Language and Cultural Diversity in ADM: Australia in the Asia Pacific

PROJECT SUMMARY

Image by Milan Boie

Language and Cultural Diversity in ADM: Australia in the Asia Pacific

Focus Areas: News & Media, Mobilities, Social Services
Status: Active

This project investigates the challenges and opportunities for cultural and linguistic diversity in automated decision making (ADM) across Australia and the Asia-Pacific region. Focusing upon language and cultural diversity as the central concern, the project aims to better understand the ways in which AI and ADM may be utilised to promote diversity and social cohesion across our region, in addition to identifying the roles of bias and manipulation in ADM.

Moving beyond the dominant paradigms and voices that inform debates about contemporary technologies (e.g. Anglo-centric and superpowers-focused), the Diversity in ADM project focuses on: (1) culturally and linguistically diverse (CALD) communities in Australia, and (2) communities across the Asia-Pacific. It has significance in highlighting both the comparative and connecting perspectives, viewpoints and life experiences, and dialogues in relation to ADM and in increasing creativity and problem-solving capabilities in our multicultural society and workplace.

PROJECT OBJECTIVES

  • Develop a better understand the landscapes of ADM across the region, including the role of NGOs, industry, government and other stakeholders; 
  • Empower community members to participate in dialogues concerning diversity in ADM; and
  • Build capacity for community organisations in collective bargaining with public policymakers for inclusive and equitable ADM policies.

MORE INFORMATION

RESEARCHERS

Haiqing Yu

Prof Haiqing Yu

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

ADM+S Chief Investigator Heather Horst

Prof Heather Horst

Project Co-Leader and Chief Investigator,
University of Western Sydney

Learn more

Deborah Lupton

Prof Deborah Lupton

Project Co-Leader and Chief Investigator,
UNSW

Learn more

ADM+S Chief Investigator Anthony McCosker

Prof Anthony McCosker

Chief Investigator,
Swinburne University
Learn more

Sarah Pink

Prof Sarah Pink

Chief Investigator,
Monash University

Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University

Learn more

Daniel Featherstone

Dr Daniel Featherstone

Research Fellow,
RMIT University
Learn more

Gerard Goggin

Prof Gerard Goggin

Associate Investigator,
University of Western Sydney
Learn more

ADM+S Associate Investigator Jenny Kennedy

Assoc Prof Jenny Kennedy

Associate Investigator,
RMIT University
Learn more

Dang Nguyen

Dr Dang Nguyen

Research Fellow,
RMIT University

Learn more

Dr Damiano Spina

Dr Damiano Spina

Associate Investigator,
RMIT University

Learn more

Yong-Bin Kang

Dr Yong-Bin Kang

Research Fellow,
Swinburne University

Learn more

Thao Phan

Dr Thao Phan

Research Fellow,
Monash University
Learn more

Xiaofang Yao

Dr Xiaofang Yao

Affiliate,
Federation University

Learn more

Prof Jason G. Karlin

Research Partner,
Interfaculty Initiative in Information Studies,
University of Tokyo, Japan

Learn more

Lee Kwang-Suk

Prof Kwang-Suk Lee

Research Partner,
Seoul National University of Science & Technology,
South Korea

Learn more

Cheryll soriano

Prof Cheryll Ruth Soriano

Research Partner,
La Salle Institute of Governance and Social Development Research Center,
De La Salle University, The Philippines

Learn more

Prof-Jack-Linchuan-Qiu 4x5small

Prof Jack Qiu

Research Partner,
Asian Communication Research Centre,
Nanyang Technological University, Singapore

Learn more

PARTNERS

Digital Asia Hub logo

Digital Asia Hub
(Hong Kong)

Visit website

Telstra

Telstra

Visit website

COLLABORATORS

Centre for Trusted
Internet and Community

Visit website

Generative Authenticity

PROJECT SUMMARY

Generative Authenticity

Focus Areas: News & Media, Mobilities, Social Services, Health
Status: Active

Authenticity is a key problem for understanding and managing the impacts of generative AI and synthetic media in society, and a central target for automated decision-making systems in the information and media environment. From trustworthy news reporting to identity verification for social services and the everyday risk of scams, generative AI and synthetic media present significant real-world implications for practitioners, institutions, and publics in Australia and elsewhere. 

A wide range of technical solutions collectively understood as authenticity infrastructure promise to address these issues; but if adopted and embedded at scale, some of these solutions could have potentially significant downstream effects on stakeholders and implications for society.

This project will critically examine the assumptions underpinning these developments and debates, assess the technical and legal challenges associated with them, and explore novel technical responses that contribute to more responsible, ethical and inclusive ADM systems.

In doing so, the project will draw on the multidisciplinary  expertise of the Centre and our partners explore authenticity as both a socio-technical challenge and as a contested cultural idea. We address these challenges in practical and experimental ways within the innovative and Generative AI Test Range environment. It will also examine what happens after any determination of authenticity, including mechanisms for explaining and communicating determinations and increasing trust in such measures.

PROJECT OBJECTIVES

  • Produce a cross-disciplinary understanding of the problem of authenticity in the context of Generative AI;
  • Study and map the field of Authenticity-as-a-Service (AaaS), providing a detailed account of its infrastructure, operations, and political economy; 
  • Analyse how the integration of authenticity infrastructure is already playing out in practice in specific sectors, and impacting or likely to impact specific communities;
  • Within the Generative AI Test Range environment, simulate and evaluate competing ADM techniques for addressing the problem of authenticity in a range of real-world scenarios; and
  • Build on our findings to develop improved tools and techniques, and produce and share guidelines for explanation and communication for a range of stakeholders and audiences.

MORE INFORMATION

RESEARCHERS

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Project Co-Leader and Chief Investigator,
QUT

Learn more

ADM+S Chief Investigator Christopher Leckie

Prof Christopher Leckie

Project Co-Leader and Chief Investigator,
University of Melbourne

Learn more

ADM+S Chief Investigator Anthony McCosker

Prof Anthony McCosker

Project Co-Leader and Chief Investigator,
Swinburne University

Learn more

Michael Richardson

Assoc Prof Michael Richardson

Project Co-Leader and Associate Investigator,
UNSW

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
University of Melbourne

Learn more

Prof Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

Jeffrey Chan

Assoc Prof Jeffrey Chan

Associate Investigator,
RMIT University

Learn more

ADM+S Investigator Sarah Erfani

Dr Sarah Erfani

Associate Investigator,
University of Melbourne

Learn more

ADM+S Investigator Timothy Graham

Assoc Prof Timothy Graham

Associate Investigator,
QUT

Learn more

Ariadna Matamoros Fernandez profile picture

Dr Ariadna Matamoros-Fernández

Associate Investigator,
QUT

Learn more

Dr Aaron Snoswell

Dr Aaron Snoswell

Associate Investigator,
QUT

Learn more

ADM+S partner investigator Wiebke Loosen

Prof Wiebke Loosen

Partner Investigator
Hans Bredow Institut

Learn more

ADM+S Investigator Julia Stoyanovich

Assoc Prof Julia Stoyanovich

Partner Investigator,
New York University

Learn more

Ned Watt

Ned Watt

PhD Student,
QUT

Learn more

Daniel Binns

Dr Daniel Binns

Affiliate,
RMIT University

Learn more

William He

Will He

Affiliate,
QUT

Learn more

Dr Silvia Montaña-Niño profile picture

Dr Silvia Montaña-Niño

Affiliate,
University of Melbourne

Learn more

Luke Munn

Dr Luke Munn

Affiliate,
The University of Queensland

Learn more

Michelle Riedlinger

Dr Michelle Riedlinger

Affiliate,
QUT

Learn more

TJ Thomson

Dr TJ Thomson

Affiliate,
RMIT University

Learn more

Kevin Witzenberger

Dr Kevin Witzenberger

Affiliate,
QUT

Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

Hans Bredow Institut Logo

Hans Bredow
Institut

Visit website

GenAISim: Simulation in the Loop for Multi-Stakeholder Interactions with Generative Agents

PROJECT SUMMARY

GenAISim: Simulation in the Loop for Multi-Stakeholder Interactions with Generative Agents

Focus Areas: News & Media, Mobilities, Social Services, Health
Status: Active

Traditional decision-making processes often struggle to adapt to the dynamic and multifaceted nature of the modern world. This research addresses a higher-level profound need for advanced automated decision-making tools that can address complex, context-rich challenges in society.

This project will investigate a hybrid decision-making system, leveraging cooperative knowledge from multiple stakeholders through socio-technical observations, and knowledge priors in Large Language Models (LLMs) and open datasets.

It will develop GenAISim, a novel suite of generative and data driven simulations, useful for depicting current and future urban scenarios, including in mobility, urban policymaking, and health domains. Through a multidisciplinary sociotechnical framework of investigation, this project will establish a new simulation in the loop paradigm.

PROJECT OBJECTIVES

  • Explore LLM agent-based synthetic data generation techniques to simulate and augment human behaviours in diverse contexts;
  • Develop a robust framework for hypothesis testing of real-world observations and relationships, while avoiding spurious correlations;
  • Investigate diverse stakeholder settings, often with nonoverlapping and potentially conflicting objectives, priorities, constraints, incentives and pain points; and
  • Explore questions around hybrid decision making – if an LLM agent is substituting for a decision maker in contexts.

MORE INFORMATION

RESEARCHERS

Prof Flora Salim

Prof Flora Salim

Project Co-Leader and Chief Investigator,
UNSW

Learn more

Hao Xue

Dr Hao Xue

Project Co-Leader and Associate Investigator,
UNSW

Learn more

Kim Weatherall

Prof Kimberlee Weatherall

Project Co-Leader and Chief Investigator,
University of Sydney

Learn more

Sarah Pink

Prof Sarah Pink

Project Co-Leader and Chief Investigator,
Monash University

Learn more

Daniel Angus

Prof Daniel Angus

Project Co-Leader and Chief Investigator,
QUT

Learn more

Jeffrey Chan

Assoc Prof Jeffrey Chan

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
University of Melbourne

Learn more

ADM+S Chief Investigator Christopher Leckie

Prof Christopher Leckie

Chief Investigator,
University of Melbourne

Learn more

ADM+S Investigator Sarah Erfani

Dr Sarah Erfani

Associate Investigator,
University of Melbourne

Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University

Learn more

Debora Lanzeni

Dr Debora Lanzeni

Associate Investigator,
Monash University

Learn more

ADM+S Chief Investigator Falk Scholer

Prof Falk Scholer

Associate Investigator,
RMIT University

Learn more

Dr Aaron Snoswell

Dr Aaron Snoswell

Associate Investigator,
QUT

Learn more

Dr Damiano Spina

Dr Daminao Spina

Associate Investigator,
RMIT University

Learn more

Maarten de Rijke

Prof Maarten de Rijke

Partner Investigator,
University of Amsterdam

Learn more

ADM+S Partner Investigator Ouri Wolfson

Prof Ouri Wolfson

Partner Investigator,
University of Illinois

Learn more

Shohreh Deldari

Dr Shohreh Deldari

Research Fellow,
UNSW

Learn more

Tiberio Caetoano

Prof Tiberio Caetano

Affiliate,
Gradient Institute

Learn more

PARTNERS

Bendigo Health logo

Bendigo
Hospital

Visit website

Halmstad University logo

Halmstad
University

Visit website

University of Amsterdam logo

University of
Amsterdam

Visit website

COLLABORATORS

Gradient Institute logo

Gradient Institute

Visit website

University of Illinois

Visit website

Critical Capabilities for Inclusive AI

PROJECT SUMMARY

Critical Capabilities for Inclusive AI

Focus Areas: News & Media, Mobilities, Social Services, Health
Status: Active

Inclusive AI is related to, but distinct from Responsible AI and the ethical principles and governance frameworks currently in development. At base it involves ensuring that all members of society benefit from AI tools and ADM systems and can participate in their design or respond to their deployment. We see capabilities – machine and human – as central to how inclusive AI might be achieved.

While much of the research focus is currently targeting the features, functions and ‘use cases’ of LLMs and other AI model types, not enough emphasis is placed on the ‘human factors’ or the co-learning and socialisation taking place in real-world settings and among different groups using these tools and systems.

This project addresses the knowledge, skills and literacies – the critical capabilities – needed to achieve inclusive AI in Australia. It will work with research partners, consumers and communities to better understand the capabilities and resources needed to access and use AI tools including Generative AI. Central to the project is the AI Capabilities Lab, a platform and space to experiment, observe and evaluate the use of new AI tools with our industry partners and members of the public.

Through the AI capabilities Lab and participatory research methods, the project will build an evidence base about the shifting lines of expertise, knowledge and decision making in organisational and everyday life settings as people begin to use AI tools.

PROJECT OBJECTIVES

  • Develop a model of AI capability and literacy for AI inclusion, tested with key domain areas and target populations;
  • Generate empirical evidence about the way people and organisations are using AI tools, and their potential for alleviating or deepening digital inequalities;
  • Co-design resources with partner organisations and their communities and consumers to enhance inclusive AI literacy and capability and foster responsible forms of ‘social governance’ for AI use and ADM processes; and
  • Develop and test AI capability and usage metrics, test and evaluate AI tools and systems through user studies and explore meta-evaluation approaches for targeted AI tools and applications.

MORE INFORMATION

RESEARCHERS

ADM+S Chief Investigator Anthony McCosker

Prof Anthony McCosker

Project Co-Leader and Chief Investigator,
Swinburne University

Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

Kath Albury

Prof Kath Albury

Project Co-Leader and Associate Investigator,
Swinburne University

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Chief Investigator,
QUT

Learn more

Paul Henman headshot

Prof Paul Henman

Chief Investigator,
The University of Queensland

Learn more

Profile image of Jackie Leah Scully

Prof Jackie Leach Scully

Chief Investigator,
UNSW

Learn more

Deborah Lupton

Prof Deborah Lupton

Chief Investigator,
QUT

Learn more

Haiqing Yu

Prof Haiqing Yu

Chief Investigator,
RMIT University

Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University

Learn more

James Meese

Assoc Prof James Meese

Associate Investigator,
RMIT University

Learn more

ADM+S Associate Investigator Jenny Kennedy

Assoc Prof Jenny Kennedy

Associate Investigator,
RMIT University

Learn more

Sharon Parkinson

Assoc Prof Sharon Parkinson

Associate Investigator,
Swinburne University

Learn more

Dr Damiano Spina

Dr Daminao Spina

Associate Investigator,
RMIT University

Learn more

Rowan Wilken

Assoc Prof Rowan Wilken

Associate Investigator,
RMIT University

Learn more

ADM+S Investigator Julia Stoyanovich

Assoc Prof Julia Stoyanovich

Partner Investigator,
New York University

Learn more

Daniel Featherstone

Dr Daniel Featherstone

Research Fellow,
RMIT University

Learn more

Awais Hameed Khan profile image

Dr Awais Hameed Khan

Research Fellow,
The University of Queensland

Learn more

Yong-Bin Kang

Dr Yong-Bin Kang

Research Fellow,
Swinburne University

Learn more

Amanda Lawrence

Dr Amanda Lawrence

Affiliate,
RMIT University

Learn more

TJ Thomson

Dr TJ Thomson

Affiliate,
RMIT University

Learn more

Xiaofang Yao

Dr Xiaofang Yao

Afilliate,
Federation University

Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

Australian Red Cross Logo

Australian
Red Cross

Visit website

Telstra

Telstra

Visit website

NYU Ultra Violet Logo

Centre for Responsible AI
New York University

Visit website

The Australian Search Experience 2.0

PROJECT SUMMARY

The Australian Search Experience 2.0

Focus Areas: News & Media
Status: Active

The first phase of The Australian Search Experience project launched in late July 2021 and over 12 months collected over 350 million search results from more than 1,000 participants. Results from phase 1 showed that personalisation of results for generic queries in major search engines is minimal, and generally limited to ensuring geographic relevance for users. Yet partisanship is rife. Eight billion queries are issued daily to Google alone, and many partisans use search engines for fact checking, can it really be that this ubiquitous ADM plays no role in the growing partisanship of society? 

The Australian Search Experience 2.0 will address a critically overlooked part of the story. We already know individual searchers formulate their queries very differently, and that this directly affects the results obtained, but understanding the extent and form of query variation as well as its impact is methodologically challenging.

The query is one of the key means people have to control not just search, but with the rise of generative AI, text queries and instructions are becoming the dominant way that ADM systems are controlled. 

When considering the impact of search on society, we know that individual searchers formulate their queries very differently, and that this directly affects the results obtained.

This project will generate significant new knowledge about the breadth and quality of information returned in response to diverse search queries, and offers a significantly more realistic perspective than past research for the impact of variations in users’ queries on the results recommended to them. It will significantly advance the state of the art in the field by developing novel methodologies for the study of search and recommendation in conventional (text-based) and emerging (voice-based and AI assisted) search interfaces, and producing new insights into their impact on users and content creators.

Findings from this research will have both productive and preventative implications: a better understanding of query diversity for a given topic will enable the designers of textbased and non-text-based search interfaces to enhance the ability of such interfaces to produce quality results; similarly, it will also enable educators and non-profit organisations to develop more targeted interventions for improving search literacy and preventing the spread of misinformation.

PROJECT OBJECTIVES

  • Explore novel ways of measuring effectiveness of search engines, by considering a diverse pool of demographic groups that represent the breadth of the Australian population, and observing their approaches to developing search queries;
  • Evaluate the range of search results that such queries return using AI to simulate the full diversity of search queries on controversial topics;
  • Develop an understanding of how non-traditional search interfaces and contextual factors drive query variation; and
  • Determine relationships between queries and search results for specific types of query, public interest topics and key news and information sources such as Wikipedia projects.

MORE INFORMATION

RESEARCHERS

ADM+S Chief Investigator Mark Sanderson

Prof Mark Sanderson

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

Axel Bruns

Prof Axel Bruns

Project Co-Leader and Chief Investigator,
QUT

Learn more

Daniel Angus

Prof Daniel Angus

Chief Investigator,
QUT

Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University

Learn more

James Meese

Assoc Prof James Meese

Associate Investigator,
RMIT University

Learn more

Ariadna Matamoros Fernandez profile picture

Dr Ariadna Matamoros-Fernández

Associate Investigator,
QUT

Learn more

ADM+S Chief Investigator Falk Scholer

Prof Falk Scholer

Associate Investigator,
RMIT University

Learn more

Dr Damiano Spina

Dr Damiano Spina

Associate Investigator,
RMIT University

Learn more

Johanne Trippas

Dr Johanne Trippas

Associate Investigator,
RMIT University

Maarten de Rijke

Prof Maarten de Rijke

Partner Investigator,
University of Amsterdam

Learn more

ADM+S professional staff Abdul Obeid

Dr Abdul Obeid

Data Engineer,
QUT

Learn more

Amanda Lawrence

Dr Amanda Lawrence

Affiliate,
RMIT University

Learn more

Natali Helberger

Prof Natali Helberger

Research Partner,
University of Amsterdam

Learn more

Judith Moller

Prof Judith Möller

Research Partner,
Hans Bredow Institut

Learn more

Kim Osman

Dr Kim Osman

Research Partner,
QUT

Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

AlgorithmWatch logo

AlgorithmWatch

Visit website

Hans Bredow Institut Logo

Hans-Bredow-Institut

Visit website

University of Amsterdam logo

University of Amsterdam

Visit website

Evaluating Automated Cultural Curating and Ranking Systems with Synthetic Data

PROJECT SUMMARY

Evaluating Automated Cultural Curating and Ranking Systems with Synthetic Data

Focus Areas: News & Media
Status: Active

As culture becomes increasingly mediated through automated systems, it is crucial to devise means of evaluating their performance from a societal perspective. Automated systems increasingly navigate a sea of content on our behalf, shaping the flow of culture and cultural products such as news, information retrieval and search, recommender systems and music and video.

This project addresses the question of how to study and evaluate the role that automated systems play in shaping this flow. It brings together technical innovations in the use of synthetic data for hybrid approaches to simulate a broad range of behaviours that influence automated decision-making systems, with a specific focus on ranking-based systems.

It will build tools and approaches to evaluate how recommender and ranking-based systems might incorporate cultural and civic values, such as better representation of artists of colour on streaming platforms or higher visibility and circulation for information of significant public interest.

This project will develop strategies for intervening in the automated flow of culture to advance Centre priorities of responsive, ethical, and inclusive automated decision making.

PROJECT OBJECTIVES

  • Understand what cultural products people are exposed to and in what order and why;
  • Evaluate whether and how ranking-based systems might incorporate factors and values that range from commercial to cultural and civic;
  • Understand how different automated systems create an overall cultural milieu through their combinations and interactions with one another;
  • Provide key insights into how ADMs are experienced in specific sites;
  • Develop ways to ensure that automated intervention in cultural flows aligns with social priorities, including commitments to diversity, fairness, and inclusion; and
  • Demonstrate the appropriate contexts and methods for using synthetic data in evaluating ADM systems.

MORE INFORMATION

RESEARCHERS

Mark Andrejevic

Prof Mark Andrejevic

Project Co-Leader and Chief Investigator,
Monash University

Learn more

Jeffrey Chan

Assoc Prof Jeffrey Chan

Project Co-Leader and Associate Investigator,
RMIT University

Learn more

Kylie Pappalardo profile picture

Dr Kylie Pappalardo

Project Co-Leader and Associate Investigator,
QUT

Learn more

James Meese

Assoc Prof James Meese

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Chief Investigator,
QUT

Learn more

Prof Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

ADM+S Chief Investigator Mark Sanderson

Prof Mark Sanderson

Chief Investigator,
RMIT University
Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University
Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University
Learn more

Joel Stern

Dr Joel Stern

Associate Investigator,
RMIT University
Learn more

Patrik Wikstrom

Prof Patrik Wikstrom

Associate Investigator,
QUT

Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

NYU Ultra Violet Logo

Centre for Responsible AI
New York University

Visit website

ADM, Ecosystems and Multispecies Relationships

PROJECT SUMMARY

Person standing in front of large digital screen with landscape image

ADM, Ecosystems and Multispecies Relationships

Focus Areas: News & Media, Mobilities, Social Services, Health
Status: Active

Automated Decision-Making (ADM) has become increasingly implicated in the relationships between people and other species and ecosystems. From delivery drones to digital bioacoustics, smart farming, smart garbage trucks to conservation and computation, proliferating ADM-enabled technologies are situated within and interact in complex ways with both social and eco-systems to create new mediations between humans, technologies, animals, and environments with diverse and unexpected consequences.

This project will make an innovative and transformational contribution to the advancement of knowledge about the impacts and entanglements of ADM with ecosystems and the capacity of institutions to make responsible decisions about ADM implementations, practices, and assessments.

 Drawing on interdisciplinary socio-technical research practices, researchers will undertake an inclusive approach that brings together diverse knowledges, methods, and sites. In collaboration with partners and communities this project will produce the ADM+Ecosystem Playbook, a policy and practice tool kit that includes addressing the potential for an environmental impact assessment legislative, policy and standards framework for ADM in Australia.

This project will intervene in the ongoing debates about ‘safe and responsible AI’ to critically examine the ecosystem impacts of ADM/AI and prioritise sustainable futures that benefit society and more-than-human ecologies alike.

PROJECT OBJECTIVES

  • To deliver an original account of how entanglements between ADM systems, diverse human stakeholder groups, other non-human species and Australian ecosystems/environments are evolving, with particular attention to Australia’s unique exposure to climate extremes of heat, drought, flood, and fire and demands of automated technologies to cover distance;
  • To generate new experimental and arts practice based methodologies for investigating, representing and creating public and diverse stakeholder engagement with the relationship between humans, other species and ADM systems, including challenges of environmentally responsible ADM; and
  • To produce accessible, practical recommendations for policies and standards that enable industry, government, civil society and advocacy organisations to apply a responsible and sustainable approach to relations between ADM systems, eco systems and other species, with the aim of intervening in the discourse, conception, and implementation of ‘safe and responsible AI’ and the wider public and civil society understanding of ADM and its impacts.

MORE INFORMATION

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Project Co-Leader and Chief Investigator,
University of Melbourne

Learn more

Deborah Lupton

Prof Deborah Lupton

Project Co-Leader and Chief Investigator,
UNSW

Learn more

Michael Richardson

Assoc Prof Michael Richardson

Project Co-Leader and Associate Investigator,
UNSW

Learn more

Sarah Pink

Prof Sarah Pink

Project Co-Leader and Chief Investigator,
Monash University

Learn more

Yolande Strengers

Prof Yolande Strengers

Project Co-leader and
Associate Investigator,
Monash University
Learn more

Loup Cellard

Dr Loup Cellard

Affiliate,
Datactivist/Sciences Partner Organisation
Learn more

ADM+S Investigator Fiona Haines

Prof Fiona Haines

Affiliate,
University of Melbourne

Learn more

Luke Munn

Dr Luke Munn

Affiliate,
University of Queensland
Learn more

James Parker

Assoc Prof James Parker

Affiliate,
University of Melbourne

Learn more

PARTNERS

Consumer Policy Research Centre Logo

Consumer Policy
Research Centre

Visit website

COLLABORATORS

CHOICE

Visit website

Australian Ad Observatory: Investigating mobile and dynamic advertising via computational and participatory approaches

PROJECT SUMMARY

Australian Ad Observatory: Investigating mobile and dynamic advertising via computational and participatory approaches

Focus Areas: News & Media
Status: Active

Advertising remains the dominant model for supporting commercial media platforms, and continues to pioneer strategies of data driven customisation and targeting. Advertisers are at the forefront of experimenting with automated digital media across recommendation, targeting, synthetic and augmented content, logistics and retail. Revenue from advertising funds the digital media platforms that in turn invest in engineering automated models that curate, augment and synthesise our media experience. 

Phase one of the Ad Observatory pioneered a way to observe the targeting of social media advertising across populations of users. We generated the largest known collection of targeted ads that people encounter on Facebook in Australia – 328,107 unique ads from 1909 participants –and built world-first research infrastructure that involved citizens in doing so. The project led to significant findings and impact, and new ways of approaching the study of automated advertising, not only in terms of individually targeted, discrete ads, but as ongoing sequences of ads that are ‘tuned’ to work in tandem with people’s identities and daily rhythms.

Responding to significant recent and ongoing developments in automated advertising (including Generative AI), the Phase 2 Ad Observatory will develop approaches for studying contemporary media and information environments, where there are no longer either shared flows of content, nor stable texts.

It will conduct participatory research with diverse groups of Australians to provide visibility into the targeting of harmful products to particular groups. It will explore experiences of advertising and understand its cultural impact, combining citizen science with data collection.

This project will significantly advance our conceptual understanding of automated advertising, playing a crucial role in documenting the emergence of this new form of advertising and enabling industry, civil society and government to respond to the challenges it will create to observability and accountability.

PROJECT OBJECTIVES

  • Develop research infrastructure and methods to observe dark, ephemeral, and automatically generated ad content and sequences, using a prototype Mobile Ad Observatory Toolkit to collect digital ads across platforms;
  • Conduct participatory research with diverse groups of Australians to explore their experiences of advertising and understand its cultural impact, combining citizen science with data collection;
  • Develop tools for automatically identifying defined categories of advertising (such as political ads and ads for harmful products and services);
  • Provide visibility into the targeting of political advertising during the next Federal election in Australia by recruiting undecided voters in marginal and swinging electorates;
  • Develop a detailed account of competing advertising explanatory models: those offered by the platforms, those offered by users, and those informed from our observations;
  • Conceptualise and develop models for simulating automated advertising at individual and cultural levels, creating tools for users to compare ad sequences and question the automated processes behind them;
  • Examine implications of automated advertising, including the use of generative AI, for current and contemplated platform governance policies, legal tools, and regulatory frameworks; and
  • Collaborate with partners to build a public interest network focused on digital advertising, promoting observability, advocacy, and influencing the development of responsible advertising practices.

MORE INFORMATION

  • Develop research infrastructure and methods to observe dark, ephemeral, and automatically generated ad content and sequences, using a prototype Mobile Ad Observatory Toolkit to collect digital ads across platforms;
  • Conduct participatory research with diverse groups of Australians to explore their experiences of advertising and understand its cultural impact, combining citizen science with data collection;
  • Develop tools for automatically identifying defined categories of advertising (such as political ads and ads for harmful products and services);
  • Provide visibility into the targeting of political advertising during the next Federal election in Australia by recruiting undecided voters in marginal and swinging electorates;
  • Develop a detailed account of competing advertising explanatory models: those offered by the platforms, those offered by users, and those informed from our observations;
  • Conceptualise and develop models for simulating automated advertising at individual and cultural levels, creating tools for users to compare ad sequences and question the automated processes behind them;
  • Examine implications of automated advertising, including the use of generative AI, for current and contemplated platform governance policies, legal tools, and regulatory frameworks; and
  • Collaborate with partners to build a public interest network focused on digital advertising, promoting observability, advocacy, and influencing the development of responsible advertising practices.

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Project Co-Leader and Chief Investigator,
University of Melbourne

Learn more

Nicholas Carah

Assoc Prof Nicholas Carah

Project Co-Leader and Associate Investigator,
University of Queensland

Learn more

Mark Andrejevic

Prof Mark Andrejevic

Project Co-Leader and Chief Investigator,
Monash University

Learn more

Daniel Angus

Prof Daniel Angus

Project Co-Leader and Chief Investigator,
QUT

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Project Co-Leader and Chief Investigator,
QUT
Learn more

Prof Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

Kim Weatherall

Prof Kimberlee Weatherall

Chief Investigator,
University of Sydney
Learn more

ADM+S Investigator Timothy Graham

Assoc Prof Timothy Graham

Associate Investigator,
QUT

Learn more

ADM+S professional staff Abdul Obeid

Dr Abdul Obeid

Data Engineer,
QUT
Learn more

Thao Phan

Dr Thao Phan

Research Fellow,
Monash University
Learn more

Maria-Gemma Brown research partner

Maria-Gemma Brown

PhD Student,
The University of Queensland

Learn more

Lauren Hayden

Lauren Hayden

PhD Student,
The University of Queensland
Learn more

César Albarrán-Torres

Dr César Albarrán-Torres

Affiliate,
Swinburne University

Learn more

Robbie Fordyce

Dr Robbie Fordyce

Affiliate,
Monash University

Learn more

Giselle Newton

Dr Giselle Newton

Affiliate,
The University of Queensland

Learn more

Verity Trott

Dr Verity Trott

Affiliate,
Monash University
Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

Consumer Policy Research Centre Logo

Consumer Policy
Research Centre

Visit website

COLLABORATORS

CHOICE

Visit website

FARE Logo

Foundation for Alcohol
Research & Education

Visit website

Vic Health Logo

VicHealth

Visit website

Automated informality: generative frictions in ADM systems

PROJECT SUMMARY

A seller stands next to his electronic devices outside in front of graffiti art on wall.

Automated informality: generative frictions in ADM systems

Focus Areas: News & Media
Research Program: People, Data, Machines & Institutions
Status: Active

Informality, especially in economic practice, poses a recurrent problem in development literature. Economic informality is broadly associated with weaker economic outcomes: countries with larger informal sectors have lower per capita incomes, greater poverty, less financial development, and weaker growth in output, investment, and productivity. As such regimes across the globe have sought to intervene in, and formalize the informal sector through worker registration drives, technology transfers, and other interventions which attempt to expand the reach of the formal economy bringing swaths of the working population under regimes of taxation, workplace safety, and enhanced productivity.

Recently, such interventions have turned on the possibilities and promises of automation. While industrial robotics systems boost manufacturing productivity, digital platforms make possible immediate and traceable circulation of funds, even as biometric databases enable automated identity verification in commercial and civic contexts.  Here new technologies of automation hold out the potential to formalize economic practices by extending standardized protocols in the form of apps, database architectures, and machinery.

Scholars of informal work have emphasized that informal and formal economic practices have long been intertwined, and they are connected by exchanges of personnel, ideas, content, and capital as highly contingent interactions. Especially in the Global South, the informal is not exceptional but typical with informality characterizing most economic practices. In India, for example, the rise of formal IT outsourcing firms has been matched by the growth of temporary and unregulated service workers who clean the offices, fix the meals, and provide transportation to professional employees.

In Brazil, wageless trash collectors sort recyclable items from Rio de Janeiro’s municipal waste dumps enabling the operation of this public infrastructure while extracting a livelihood from reselling this waste. Far from eliminating informal economies contemporary regimes of accumulation generate value by weaving formal and informal practices together.

Currently missing from this body of scholarship is a range of contingent and non-standard work that proliferates as a result of the friction that exists within automated systems as complex self-coordinating and self-organising mechanisms. This type of work – which we call small automation – is different from gig work in that it is unregulated, opportunistic, and marginalised; it is largely invisible and opaque, but unlike ghost work, its invisibility is key to its survival.

Small automation is different from both gig work and ghost work in the sense that it encompasses a range of informal enterprises created by informal actors that circumvent, exploit, or co-opt automated systems, rather than being deployed by Silicon Valley to develop new technologies.

This project maps a range of informal automated activities that proliferate within automated systems across various empirical domains, such as click farming, CAPTCHA hacking, phone farming, dropshipping, OTP scams, fraudulent loan apps, and free jacking. The proliferation of automated informality can create unexpected implications for the operation of automated systems and our information environment more generally. Our focus on mapping automated informality works to supplement current research on gig work and ghost work while demonstrating the theoretical and empirical value of examining automated systems in context.

RESEARCHERS

Dang Nguyen

Dr Dang Nguyen

Lead Investigator,
RMIT University

Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT

Learn more

Rakesh Kumar

Rakesh Kumar

PhD Student,
Western Sydney University

Learn more

AI ReWired: How communities are using AI to Support Social and Environmental Justice

PROJECT SUMMARY

A persons neck wearing a microship necklace

AI ReWired: How communities are using AI to Support Social and Environmental Justice

Focus Areas: News & Media, Mobilities, Health
Research Program: People
Status: Completed

The future we are being sold is an automated wonderland, a techtopia that will use algorithms to heal our ecological crisis and restore social justice. A dream world where we enjoy endless innovation and growth in sparkling smart cities, where we are liberated from the burden of work, where the future of our species lies in billionaire funded missions to Mars.

But what if this promise sounds more like a nightmare?
What are the alternatives?

The AI ReWired project uses co-creative documentary film practice to uncover how diverse communities utilise AI systems to protect the environment, support social justice and promote fairness in their communities.

RESEARCHERS

Jeni Lee

Jeni Lee

Lead Investigator,
Monash University

Learn more

ADM+S Chief Investigator Sarah Pink

Prof Sarah Pink

Chief Investigator,
Monash University

Learn more

Dr Damiano Spina

Dr Damiano Spina

Associate Investigator,
RMIT University

Learn more

Yolande Strengers

Prof Yolande Strengers

Associate Investigator,
Monash University

Learn more

Georgia Van Toorn

Dr Georgia van Toorn

Associate Investigator,
UNSW

Learn more

Thao Phan

Dr Thao Phan

Research Fellow,
Monash University

Learn more

Emma Quilty

Dr Emma Quilty

Affiliate,
Monash University

Learn more

Melissa Gregg

De Mel Gregg

Senior Industry Fellow,
RMIT

Learn more

Nonie May

Dr Nonie May

Project support,
Monash University

Learn more

Humans, Machines, and Decision Responsibility

PROJECT SUMMARY

Businessman using cell phone on subway train

Humans, Machines, and Decision Responsibility

Focus Areas: News & Media, Social Services, Mobilities, Health
Research Program: Institutions, Machines
Status: Active

Automated decision-making provokes a range of anxieties around transparency, equality, and accountability. A key response has been the call to ‘re-humanise’ automated decisions, with the hope that human control of automated systems might defend human values from mindless technocracy. Regulation of automated decision-making and AI often embeds this form of human centrism by prescribing a ‘human in the loop’ and the need for automated decisions to be ‘explained’. These requirements are central elements of the risk-based approaches AI regulation currently in development.

Despite their intuitive appeal, empirical research is revealing the limitations and complexities of these approaches. AI explanations sometimes provide little that is useful for decision subjects or decision makers, and risk distracting from more meaningful interrogation of why decisions are made. A human in the loop sometimes functions as a rubber stamp for automated decisions, cleaving accountability away from the true sites of decision responsibility.

This project seeks to generate better understandings of the functions, capacities, and normative role of humans within automated decision systems. It will investigate the ways that automated systems ought to explain or be explained to humans within decision processes, and how elements of decision-making including processes, responsibility, authority, and what counts as a decision itself, are fragmented and redistributed between humans, machines, and organisations. The goal is to generate empirical knowledge of how automated systems, humans, and organisations interact in different contexts when making decisions, and to move past outdated understandings of decisions-making that are hindering effective governance of automation in new decision contexts.

RESEARCHERS

Jake Goldenfein

Dr Jake Goldenfein

Lead Investigator,
University of Melbourne

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Chief Investigator,
QUT

Learn more

Paul Henman headshot

Prof Paul Henman

Chief Investigator,
University of Queensland

Learn more

ADM+S Chief Investigator Christopher Leckie

Prof Chris Leckie

Chief Investigator,
University of Melbourne

Learn more

Prof Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University

Learn more

Kim Weatherall

Prof Kim Weatherall

Chief Investigator,
University of Sydney

Learn more

Henry Fraser

Dr Henry Fraser

Research Fellow,
QUT

Learn more

Awais Hameed Khan profile image

Dr Awais Hameed Khan

Research Fellow,
UQ

Learn more

Fan Yang

Dr Fan Yang

Research Fellow,
University of Melbourne

Learn more

Libby Young

Libby Young

PhD Student
University of Sydney

Learn more

Joe Brailsford

Joe Brailsford

Affiliate
University of Melbourne

Learn more

Fabio Mattioli

Dr Fabio Mattioli

Affiliate
University of Melbourne

Learn more

Christopher O'Neill

Dr Chris O’Neil

Affiliate,
Deakin University

Learn more

Ash Watson

Dr Ash Watson

Affiliate,
UNSW

Learn more

Australian Digital Inclusion Index

PROJECT SUMMARY

Australian Digital Inclusion Index

Focus Areas: News & Media, Social Services, Mobilities, Health
Status: Active

Digital inclusion is about ensuring that all Australians can access and use digital technologies effectively. We are experiencing an accelerating digital transformation in many aspects of economic and social life. Our premise is that everyone should have the opportunity to benefit from digital technologies: to manage their health, access education and services, participate in cultural activities, organise their finances, follow news and media, and connect with family, friends, and the wider world.

The Australian Digital Inclusion Index (ADII or “Index”) uses survey data to measure digital inclusion across three dimensions of Access, Affordability and Digital Ability. We explore how these dimensions vary across the country and across different social groups.

In partnership with Telstra and through biennial data collections presented through reports and data visualisation dashboards, the ADII is capturing and communicating the evolving state of digital inclusion in Australia. This is complemented by aligned sub-projects with local, state and federal government departments and community partners to drill down into specific digital inclusion challenges for social groups or geographical regions of interest.

A detailed measure of digital inclusion for Australia allows us to identify the critical barriers to inclusion. These may be related to accessing networks, the costs of devices or data, or skills and literacies. Through these measures, the Index shapes digital equity policy and initiatives, research, and practice to increase digital inclusion in Australia.

Visit the ADII website 

MORE INFORMATION

The Australian Digital Inclusion Index uses data from the ADM+S project, Mapping the Digital Gap. Learn more from the project brief below.

PUBLICATIONS

Uncovering digital divide in the western parkland city

Uncovering the digital divide in the Western Parkland City

ADM+S, Telstra, NSW Government, Sydney’s Parkland Councils

Report

Measuring Australia’s Digital Divide: 2023 Australian Digital Inclusion Index

ADM+S and Telstra

Report

Telstra Connected Students: Lessons for Digital Inclusion, 2022

ADM+S and Telstra

Report

Australian Digital Inclusion Index: Measuring Digital Inclusion in North-East Victorian SMEs Summary Findings Brief, 2022

Thomas, J., Parkinson, S., et al.

Report

2021 Digital Inclusion Index

ADM+S and Telstra

Report

RESEARCHERS

ADM+S Chief Investigator Anthony McCosker

Prof Anthony McCosker

Chief Investigator,
Swinburne University

Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University

Learn more

Jenny Kennedy

Assoc Prof Jenny Kennedy

Associate Investigator,
RMIT University

Learn more

Sharon Parkinson

Dr Sharon Parkinson

Associate Investigator,
Swinburne University

Learn more

Daniel Featherstone

Dr Daniel Featherstone

Senior Research Fellow,
RMIT University

Learn more

Lyndon Ormond-Parker

Assoc Prof Lyndon Ormond-Parker

Senior Research Fellow,
RMIT University

Learn more

Kieran Hegarty

Dr Kieran Hegarty

Research Fellow,
RMIT University

Learn more

RESEARCH SUPPORT

Lucy Valenta profile image

Lucy Valenta

Research Coordinator,
RMIT University

Learn more

PARTNERS

Telstra

Telstra

Visit website

Is Pricing Discriminatory: Testing Automated Decision-Making Systems in Online Insurance Markets

PROJECT SUMMARY

man and women working on laptop together

Is Pricing Discriminatory: Testing Automated Decision-Making Systems in Online Insurance Markets

Focus Areas: News & Media, Social Services, Mobilities, Health
Research Program: Data
Status: Active

Advances in data-driven and AI systems are driving significant transformation in the emerging insurance technology (insurtech) sector.

This project investigates the extent to which automated decision-making systems impact the provision of consumer insurance via pricing algorithms which may produce unfair outcomes for particular subsets of society by engaging in proxy and price discrimination.

RESEARCHERS

Kelly Lewis

Dr Kelly Lewis

Lead Investigator,
Monash University

Learn more

Mark Andrejevic

Prof Mark Andrejevic

Chief Investigator,
Monash University

Learn more

Daniel Angus

Prof Daniel Angus

Chief Investigator,
QUT

Learn more

Kim Weatherall

Prof Kim Weatherall

Chief Investigator,
University of Sydney

Learn more

Zofia Bednarz

Dr Zofia Bednarz

Associate Investigator,
University of Sydney

Learn more

Jathan Sadowski

Dr Jathan Sadowski

Associate Investigator,
Monash University

Learn more

ADM+S professional staff Abdul Obeid

Dr Abdul Obeid

Data Engineer,
QUT

Learn more

PARTNERS

CHOICE

Visit website

Consumer Policy Research Centre Logo

Consumer Policy Research Centre

Visit website

Automation and Public Space

PROJECT SUMMARY

LiDAR sensing concept

Automation and Public Space

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Program: Data
Status:
 Active

From delivery drones to digital twins to crowd surveillance, automated decision-making technologies and practices are increasingly impacting public and shared space. This project investigates how automated decision-making systems impact public and shared space via sensors that produce actionable digital simulations, artefacts, and interfaces. Through a mixed methods approach, it will examine current and potential effects of automated decision-making on the form, use, and experience of public space.

Technological development in this area is undergoing rapid change. Delivery via autonomous drone requires sensor-driven navigation systems, but the data and models they produce about public space will likely lead to modulations of that space in response. In urban and environmental governance, ‘digital twins’ are increasingly to monitor environments in real-time, simulate the impact of potential changes, and even implement those changes directly. Technologies such as these are not only increasingly deployed in Australia, but are also invented, designed, and tested here too, often in proximity to defence and defence industries.

Understanding how tools of automated spatiality reconfigure environments and the role of policy and industry in their innovation and uptake will generate new knowledge about a critical point of convergence between public space, technology, defence, and industry with national significance, as well as implications for international jurisdictions facing similar changes and challenges.

Over 3 years commencing in 2022, the project aims to answer the following questions:
• How is space-making automated across different technologies and contexts? What logics, techniques and practices are shared? What are distinct to different contexts?
• How does automated spatiality lead to the reconfiguring of public space?
• How are digital infrastructures, such as unmanned traffic management systems for civilian airspace, imagined, organised, and regulated?
• How do policy settings, industrial demands, and defence priorities shape the development and application of technologies of automated spatiality?

PUBLICATIONS

Andrejevic, M.

Journal article

Biometric Re-bordering: Environmental Control During Pandemic Times, 2022

Andrejevic, M., Volcic, Z.

Journal article

Seeing Like a Border, 2021

Andrejevic, M., Volcic, Z.

Journal article

RESEARCHERS

Michael Richardson

Assoc Prof Michael Richardson

Lead Investigator,
UNSW

Learn more

Mark Andrejevic

Prof Mark Andrejevic

Chief Investigator,
Monash University

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
University of Melbourne

Learn more

ADM+S Chief Investigator Anthony McCosker

Prof Anthony McCosker

Chief Investigator,
Swinburne University

Learn more

Jathan Sadowski

Dr Jathan Sadowski

Associate Investigator,
Monash University

Learn more

Rowan Wilken

Assoc Prof Rowan Wilken

Associate Investigator,
RMIT University

Learn more

Kelly Lewis

Dr Kelly Lewis

Research Fellow,
Monash University

Learn more

Thao Phan

Dr Thao Phan

Research Fellow,
Monash University

Learn more

Zoe Horn

Zoe Horn

Student,
Western Sydney University

Learn more

Lauren Kelly

Lauren Kelly

Student,
RMIT University

Learn more

Andrew Brooks

Dr Andrew Brooks

Affiliate,
UNSW

Learn more

Danielle Hynes

Danielle Hynes

Affiliate,
UNSW

Learn more

Christopher O'Neill

Dr Chris O’Neill

Affiliate,
Deakin University

Learn more

PARTNERS

OVIC Logo

Office of the Victorian Information Commissioner

Learn more

Risk, Rule-setters and Rule-takers: Regulatory approaches to risk in AI-supported and AI-automated decision-making for general welfare

PROJECT SUMMARY

Crowd in motion in busy precinct

Risk, Rule-setters and Rule-takers: Regulatory approaches to risk in AI-supported and AI-automated decision-making for general welfare

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Programs: Institutions
Status: Completed

This project seeks to scope several approaches to deal with Automated Decision-Making and Decision-Support Systems-Related Risks (ADM/DSS RR) through norms and provide an evaluation of those approaches for their consideration in regulatory contexts.

The standpoint is to look at risk control of those systems beyond ethics or social principles and focus the discussion on the possible interventions from the regulator’s perspective.

The overarching questions and sub-questions guiding this project are:

  • What is risk in an ADM / DS System?
    – Is it possible to define it?
    – How is it different from technological risk?
    – How is it different from the concept of risk in
    management?
    – Are all “potential harms” risks of and ADM/DSS?
    – Is there a concept of risk usable for regulatory
    purposes?
  • What types of risks are common and which ones specific to ADM/DSS?
    – Due to the nature of the risk?
    – Due to the scale of the risk?
  • What is an acceptable risk:
    – From the point of view of society as a whole
    – From the point of view of the most vulnerable groups
    – From the point of view of the owner of the AI system
    – From the point of view of the users of the system
  • Can risk be separated from questions of liability/ responsibility or do they need to be regulated together?

PUBLICATIONS

Acceptable risks in Europe’s Proposed AI Act: Reasonableness and other principles for deciding how much risk management is enough, 2023

Bello y Villarino, J.M., Fraser, H.

Journal article

The Tale of Two Automated States: Why one-size-fits-all approach to administrative law reform to accommodate AI will fail, 2023

Bello y Villarino, J.M.

Book chapter

International Human Rights, Artificial Intelligence, and the Challenge for the Pondering State: Time to Regulate? 2022

Bello y Villarino, J.M., et al.

Journal article

Legal Issues Around Autonomous Systems – Civil Liability, Fault and System Safety, 2022

Fraser, H.

Conference paper

AI Opacity and Explainability in Tort Litigation, 2022

Snoswell, A., Fraser, H., Simcock, R.

Conference paper

Where residual risks reside: A comparative approach to art 9(4) of the EU’s Proposed AI Regulation, 2021

Bello y Villarino, J.M, Fraser, H.

Working paper

RESEARCHERS

Kimberlee Weatherall

Prof Kimberlee Weatherall

Chief Investigator,
University of Sydney

Learn more

ADM+S Chief Investigator Nic Suzor

Prof Nicolas Suzor

Chief Investigator,
QUT

Learn more

José-Miguel Bello y Villarino

Dr José-Miguel Bello Villarino

Research Fellow,
University of Sydney

Learn more

Henry Fraser

Dr Henry Fraser

Research Fellow,
QUT

Learn more

PARTNERS

Gradient Institute logo

Gradient Institute

Learn more

The Toxicity Scalpel: Prototyping and evaluating methods to remove harmful generative capability from foundation models

PROJECT SUMMARY

Person with colourful text overlay

The Toxicity Scalpel: Prototyping and evaluating methods to remove harmful generative capability from foundation models

Focus Areas: News and Media
Research Programs: Machines
Status: Completed

AI language models have made significant strides over the past few years. Computers are now capable of writing poetry and computer code, producing human-like text, summarising documents, engaging in natural conversation about a variety of topics, solving math problems, and translating between languages.

This rapid progress has been made possible by a trend in AI development where one general ‘foundational’ model is developed (usually using a large dataset from the internet) and then adapted many times to fit diverse applications, rather than beginning from scratch each time.

This method of ADM development can appear time and cost effective, but ‘bakes in’ negative tendencies like the creation of toxic content, misogyny, or hate speech at the foundational layer, which subsequently spread to each downstream application.

The goal of this project is to examine how language models used in ADM systems might be improved by making modifications at the foundation model stage, rather than at the application level, where computational interventions, social responsibility, and legal liability have historically focussed.

PUBLICATIONS

First page of Journal Article: Measuring Misogyny in Natural Language Generation: Preliminary Results from a Case Study on two Reddit Communities

Measuring Misogyny in Natural Language Generation: Preliminary Results from a Case Study on two Reddit Communities,2023

Snoswell, A., Nelson, L., Xue, H., Salim, F., Suzor, N., & Burgess, J.

Journal article

RESEARCHERS

ADM+S Investigator Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

ADM+S Chief Investigator Nic Suzor

Prof Nic Suzor

Chief Investigator,
QUT

Learn more

Dr Aaron Snoswell

Dr Aaron Snoswell

Associate Investigator,
QUT

Learn more

Hao Xue

Dr Hao Xue

Associate Investigator,
UNSW

Learn more

Lucinda Nelson

Lucinda Nelson

PhD Student,
QUT

Learn more

Assessing Prospective Harms (vs Benefits) associated with ADM

PROJECT SUMMARY

Two people looking at computer screens

Assessing Prospective Harms (vs Benefits) associated with ADM

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Programs: Data, Machines, Institutions
Status: 
Completed

The project (which is now completed) was set up as a preliminary exercise in assessing prospective harms vs prospective benefits associated with ADM as a first step to amelioration. It took a two-pronged approach: firstly, focusing on individual and social harms/costs that may be associated with automated or semi-automated data processing (including collection, retention, dissemination, and other uses of data) – versus prospective benefits; and secondly, assessing the levels of risk of these harms ranging from nebulous to very significant (and acknowledging there may be
uncertain outcomes and uneven distributions). The overall aim was thus to have a fuller appreciation of harms and risks as a precursor to thinking practically about amelioration/mitigation of costs.

More specifically, the project was geared to questions of elaborating and understanding the range of prospective harms associated with loss of control over data processing for individuals, groups and society, and indeed the entirety of the living world, as a first step to finding solutions such as changes in law, or social practices, or business methods, or technologies (or some combination of these).

The principal activity of the project was to have a series of workshops planned, organised and hosted by the coordinators CI Richardson, AI Roberts and Postdoc Jiménez (with administrator Astari.Kusumawardani providing support). The workshops featured the work of diverse ADM+S CIs, AIs, Researchers and Affiliates and adopted an intense mode of interrogation and discussion along with suggestions. The aim was to assist ADMS personnel with the preparation of reports, books and scholarly articles (as well as share insights and ideas).

Topics and presenters in the workshop series included the following:
•March: Aitor Jiménez (Megan Richardson chair), Crimes of digital capitalism
•March: Ariadna Matamoros- Fernández, Rosalie Gillett, Anjalee de Silva (Aitor Jiménez chair), •Gendered harm
•April: José-Miguel Bello Villarino, Henry Fraser (Megan Richardson chair), Where residual risks reside: a comparative approach to AI risk management under the EU’s AI Act Proposal
•April: Jake Goldenfein (Megan Richardsonchair) How competing constructions of humans legitimize online advertising
•May: Simon Coghlan, Christine Parker (Andy Roberts, chair), A preliminary framework for understanding how ADM/AI technologies can harm non-human animals
•June: Lisa Archbold (Andy Roberts chair), Children’s developmental privacy
•July: Frank Pasquale/Jeannie Paterson (Megan Richardson chair: co-hosted with CAIDE), Automated grace: toward more humane benefits administration via AI
•August: James Meese (Megan Richardson chair), Regulating news recommendation: looking beyond harm
•September: Megan Richardson (Jeannie Paterson chair – co-hosted with CAIDE), Trust norms and data rights
•October: Ariadna Matamoros- Fernández, Louisa Bartolo, Luke Troynar (Aitor Jiménez chair), Addressing harmful humour as an online safety issue
•November: Damian Clifford (Megan Richardson chair), Data protection and (in)accuracy

PUBLICATIONS

Harm to Nonhuman Animals from AI: a Systematic Account and Framework, 2023

Parker, C., Coghlan, S.

Journal article

Humour as an online safety issue: Exploring solutions to help platforms better address this form of expression, 2023

Matamoros-Fernández, A., Bartolo, L., Troynar, L.

Journal article

The Crimes of Digital Capitalism, 2022

Jiménez, A., Oleson, J.C.

Journal article

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Lead Investigator,
University of Melbourne

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
University of Melbourne

Learn more

Kim Weatherall

Prof Kim Weatherall

Chief Investigator,
University of Sydney

Learn more

Zofia Bednarz

Dr Zofia Bednarz

Associate Investigator,
University of Sydney

Learn more

Simon Coghlan

Dr Simon Coghlan

Associate Investigator,
University of Melbourne

Learn more

ADM+S Chief Investigator Andrew Kenyon

Prof Andrew Kenyon

Associate Investigator,
University of Melbourne

Learn more

Ariadna Matamoros Fernandez profile picture

Dr Ariadna Matamoros-Fernández

Associate Investigator,
QUT

Learn more

James Meese

Dr James Meese

Associate Investigator,
RMIT University

Learn more

Andrew Roberts

Prof Andrew Roberts

Associate Investigator,
University of Melbourne

Learn more

ADM+S Investigator Ivana Jurko

Ivana Jurko

Partner Investigator,
Red Cross Australia

Learn more

José-Miguel Bello y Villarino

Dr José-Miguel Bello Villarino

Research Fellow,
University of Sydney

Learn more

Anjalee de Silva

Dr Anjalee de Silva

Research Fellow,
University of Melbourne

Learn more

Henry Fraser

Dr Henry Fraser

Research Fellow,
QUT

Learn more

Dr Rosalie Gillett profile picture

Dr Rosalie Gillett

Research Fellow,
QUT

Learn more

Damian Clifford

Dr Damian Clifford

Affiliate,
ANU

Learn more

ADM+S Investigator Fiona Haines

Prof Fiona Haines

Affiliate,
University of Melbourne

Learn more

Aitor Jiménez

Dr Aitor Jiménez

Affiliate,
University of Melbourne

Learn more

Kobi Leins

Dr Kobi Leins

Affiliate,
King’s College

Learn more

Jeannie Paterson

Prof Jeannie Paterson

Affiliate,
University of Melbourne

Learn more

ADM+S Chief Investigator Megan Richardson

Prof Megan Richardson

Affiliate,
University of Melbourne

Learn more

PARTNERS

OVIC Logo

Office of the Victorian Information Commissioner (OVIC)

Learn more

Enabling digital transformation and considering digital futures within the cultural sector: Evaluating ACMI’s CEO digital mentoring project

PROJECT SUMMARY

ACMI building in the evening

Enabling digital transformation and considering digital futures within the cultural sector: Evaluating ACMI’s CEO digital mentoring project

Focus Area: News & Media
Research Program: Institutions
Status: Completed

This research investigated the enablers of digital transformation and considers digital futures within the cultural sector through evaluating the outcomes of ACMI’s CEO Digital Mentoring Program.

Funded by the Ian Potter Foundation and delivered in conjunction with the Australia Council, ACMI’s CEO Digital Mentoring Program offered strategic technology and digital mentoring for senior decision-making staff within the Australian cultural sector.

With digital platforms fundamentally reshaping how cultural content is created, distributed, and consumed, this research considered how cultural organisations might be better equipped to, and supported in, adopting, managing, and mitigating the risks associated with increasingly advanced technologies.

PUBLICATIONS

National Cultural Policy consultation: ADM+S together with the Digital Media Research Centre (DMRC), 2022

Holcombe-James, I., Pappalardo, K., et al.

Submission

From the top: learning from ACMI’s CEO Digital Mentoring Program 2021-22, 2022

Holcombe-James, I., et al.

Report

Enabling digital transformation within the cultural sector? Documenting ACMI’s CEO digital mentoring pilot program: executive summary, 2022

Holcombe-James, I.

Report

RESEARCHERS

Indigo Holcombe-James

Dr Indigo Holcombe-James

Lead Investigator

Learn more

Stephanie Livingstone

Stephanie Livingstone

PhD Student

Learn more

PARTNERS

Public Interest Litigation for AI Accountability

PROJECT SUMMARY

Woman sitting down with laptop

Public Interest Litigation for AI Accountability

Focus Areas: News and Media, Health, Social Services, Transport and Mobilities
Research Program: Institutions
Status: Active

If you have been harmed by bad automated decision-making, from robots to loan assessments, what can you do to right the wrong? What can the law do to help you? A growing number of public controversies about discriminatory, unpredictable and dangerous automated decision-making has raised questions about the most effective methods of accountability.

Through qualitative interviews with stakeholders (including class action and pro bono lawyers), this project seeks to identify the opportunities, enablers and barriers for public interest litigation to promote accountability and fairness in automated decision-making.

RESEARCHERS

ADM+S Chief Investigator Nic Suzor

Prof Nicolas Suzor

Lead Investigator,
QUT

Learn more

Henry Fraser

Dr Henry Fraser

Research Fellow,
QUT

Learn more

Zahra Stardust profile picture

Dr Zahra Stardust

Research Fellow,
QUT

Learn more

Political Economy of Sex Tech

PROJECT SUMMARY

Three hearts displayed on an LED screen

Political Economy of Sex Tech

Focus Area: News & Media
Research Programs: Data, Institutions
Status: Active

Smart sex technologies and networked apps are being used in sex and relationship education, to enhance sexual wellness and to improve sexual and reproductive health. To do so, they collect and process substantial amounts of intimate data. This project examines the political economy of ‘sex tech’ in order to identify how sexual technologies are being governed at scale, how sexual data is being collected, stored, shared and monetised, and how the material benefits of sex tech may be more equitably distributed.

It will provide empirical grounding to enrich scholarship on ethical data governance, predictive profiling and accountability of smart technologies.

PUBLICATIONS

Sex tech in an age of surveillance capitalism: Design, Data and Governance, 2024

Stardust, Z.

Book chapter

Sex tech entrepreneurs: Governing intimate data in start-up culture, 2023

Stardust, Z., Kennedy, J., Albury, K

Journal article

Surveillance does not equal safety: Police, data and consent on dating apps, 2022

Stardust, Z., Gillett, R., Albury, K.

Journal article

Public interest sex tech hackathon: speculative futures and participatory design, 2022

Stardust, Z., Kennedy, J., Albury, K.

Report

RESEARCHERS

Zahra Stardust profile picture

Dr Zahra Stardust

Lead Investigator,
QUT

Learn more

ADM+S Chief Investigator Nic Suzor

Prof Nicolas Suzor

Chief Investigator,
QUT

Learn more

Automating safety: developing better data models to help foster prosocial platforms

PROJECT SUMMARY

Blurred people crossing street

Automating safety: developing better data models to help foster prosocial platforms

Focus Area: News & Media
Research Program: Data
Status: Completed

This project identified and investigated how misunderstandings of harm and safety flow into flawed data logics and ineffective automated digital platform responses. A key output from the project included the journal article “Safety for Whom? Investigating How Platforms Frame and Perform Safety and Harm Interventions” (Gillett, Stardust, and Burgess 2023), which aimed to understand how five major digital platforms frame and define the issues of harm and safety, and identified the interventions they publicly report introducing to address these issues. Findings from this research importantly show that, in the absence of more meaningful measures that may foster safer digital cultures, at the heart of platform commitments is the scaling up of automated tools to moderate the enormous volume of content large digital platforms host.

Another key output included the journal article “Surveillance does not equal safety: Police, data and consent on dating apps” (Stardust, Gillett, and Albury, 2022). The article interrogated how dating apps focus on increasing user surveillance and other mechanisms that are deployed in the name of ‘safety’, rather than addressing the more complex questions about how their users experience and understand harm and safety, and how these understandings might better inform responses that enable users to feel safe. This paper was an extension of the article that the authors wrote for the Conversation (Gillett, Albury, Stardust, 2021).

To better understand the experiences of women online, research led by Gillett and Dr Anjalee de Silva received ADM+S seed funding for a group of the Centre’s ECRs to facilitate a Gendered online Harms Workshop. The workshop brought together a range of civil society actors and academics for whom online safety is a pressing social issue. The workshop enabled the team to better understand online harms based on users who had experienced them. Gillett and de Silva shared key findings from the research at an ADM+S Data Harms workshop. Preliminary findings were also published on the ADM+S website.

PUBLICATIONS

QUT Digital Media Research Centre submission in response to the Online Safety Bill (2021)

Prof Nicholas Suzor, Lucinda Nelson, Dr Rosalie Gillett and Prof Jean Burgess

Submission

QUT Digital Media Research Centre submission in response to the inquiry into serious vilification and hate crimes (2021)

Lucinda Nelson, Prof Nicolas Suzor, Dr Rosalie Gillett and Dr Ariadna Matamoros-Fernández

Submission

‘This is not a nice safe space’: investigating women’s safety work on Tinder (2023)

Dr Rosalie Gillett

Journal article

Safety for Whom? Investigating How Platforms Frame and Perform Safety and Harm Interventions (2022)

Prof Jean Burgess, Dr Rosalie Gillett and Dr Zahra Stardust

Journal article

‘Just a little hack’: Investigating cultures of content moderation circumvention by Facebook users (2023)

Dr Rosalie Gillett, Joanne Gray and Kaye Bondy Valdovinos

Journal article

Self-regulation and discretion (2022)

Prof Nicolas Suzor andDr Rosalie Gillett

Book chapter

Surveillance does not equal safety: Police, data and consent on dating apps (2022)

Dr Rosalie Gillett, Prof Kath Albury and Dr Zahra Stardust

Journal article

Incels on Reddit: A study in social norms and decentralised moderation (2022)

Prof Nicolas Suzor andDr Rosalie Gillett

Journal article

RESEARCHERS

Dr Rosalie Gillett profile picture

Dr Rosalie Gillett

Research Fellow,
QUT

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Lead Investigator,
QUT

Learn more

ADM+S Chief Investigator Nic Suzor

Prof Nic Suzor

Chief Investigator,
QUT

Learn more

Zahra Stardust profile picture

Dr Zahra Stardust

Research Fellow,
QUT

Learn more

Ecological Implications of Data Centres

PROJECT SUMMARY

Data centre

Ecological Implications of Data Centres

Focus Areas: News and Media, Transport and Mobility, Health, Social Services
Research Program: Institutions
Status: Active

The project seeks to understand how companies, public agencies and civil society address the environmental conditions and limitations facing the establishment and management of data centres and submarine cables in urban and coastal areas.

A central part of data centre management is heat management: servers produce heat, and as they are gathered in large numbers in close areas, temperatures rise raising the risk of fire. To overcome this, data centre operators have various techniques to cool down these facilities and avoid any risks of data loss caused by fires. Moreover, when landing, telecom subsea cables risk to damage the local biodiversity (especially marine plants).

Thus, this project will ask: what shapes the environmental impacts of data centres cooling infrastructures? What are the ecological implications involved with the landing of a telecom submarine cable or the creation of a new data centre? How are these ecological impacts made visible to stakeholders? To what extent do environmental assessments succeed in reconciling the various interests at stake (security of infrastructures, maritime trade, marine biodiversity) in the passage of a telecomunication cable? How do ecological and infrastructural vulnerabilities of both data centers and telecom submarine cables shape the world-wide interconnection of data at the heart of the digital economy?

In order to address this question, we will take as a case study the rapid growth of data centres and telecommunication subsea cables in Marseille (France), which is particularly interesting as this city is in a warm climate, making the issue of heat management more difficult there than in the north of Europe.

This project is conducted by ADM+S Research Fellow Dr Loup Cellard in collaboration with Dr Clément Marquet (Mines Paris).

PUBLICATIONS

Just Transitions in Australia: Moving Towards Low Carbon Lives Across Policy, Industry and Practice, 2022

Parker, C., Haines, F., et al.

Submission

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Lead Investigator,
University of Melbourne

Learn more

ADM+S Investigator Karen Yeung

Prof Karen Yeung

Partner Investigator,
University of Birmingham

Learn more

Loup Cellard

Dr Loup Cellard

Affiliate,
Datactivist Coop

Learn more

ADM+S Investigator Fiona Haines

Prof Fiona Haines

Affiliate,
University of Melbourne

Learn more

PARTNERS

Université de Technologie de Compiègne Logo

Université de Technologie de Compiègne

Learn more

University of Birmingham

Learn more

Platform governance of and by bots

PROJECT SUMMARY

Phone with chatbot

Platform governance of and by bots

Focus Area(s): News & Media
Research Program: Data
Status: Completed

This project brings together expertise in digital media, platform studies, and law with data science and machine learning to study the roles and data operations of bots – pre-programmed automated agents – on social media platforms. It aims to map, describe and evaluate the ways that platforms and their users make use of automated agents in governance and community management, and the competing norms and values associated with these practices.

It also examines how platforms and their communities engage in the governance of bots, including through automated moderation and technical limitations. We expect to develop new methods for the public oversight and evaluation of platform governance; as well as to understand why and how bots are understood, valued, and managed in online communities, and to suggest the implications for the benefits of bots for transparent platform governance, including by user communities.

The objectives of this project include:
Undertake a detailed empirical investigation of the role of ‘official’, sanctioned, and user-created bots in governing and managing platform cultures, and the implications of these uses of bots for equality, transparency, and user experience.

Through the data-driven analysis of a particular bot-related controversy, conduct a detailed case study of the norms attached to ‘coordination’ and ‘bot-like’ (or ‘inauthentic’) behaviour on Reddit, and how these norms are enacted and contested through community-led platform governance.
Develop new and updated frameworks for identifying and promoting the pro-social and beneficial uses of bots by platforms and their user communities.

PUBLICATIONS

Dadbot and what ‘He’ reveals about Reddit’s everyday platform culture, 2023

Carlon, D.

Conference paper

RESEARCHERS

ADM+S Investigator Timothy Graham

Dr Timothy Graham

Lead Investigator,
QUT

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Lead Investigator,
QUT

Learn more

Daniel Angus

Prof Daniel Angus

Chief Investigator,
QUT

Learn more

Axel Bruns

Prof Axel Bruns

Chief Investigator,
QUT

Learn more

ADM+S Chief Investigator Nic Suzor

Prof Nicolas Suzor

Chief Investigator,
QUT

Learn more

Transparent Machines: From Unpacking Bias to Actionable Explainability

PROJECT SUMMARY

Person typing on computer

Transparent Machines: From Unpacking Bias to Actionable Explainability

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Status: Active

ADMs, their software, algorithms, and models, are often designed as “black boxes” with little efforts placed on understanding how they work. This lack of understanding does not only impact the final users of ADMs, but also the stakeholders and the developers, who need to be accountable for the systems they are creating. This problem is often exacerbated by the inherent bias coming from the data from which the models are often trained on.

Further, the wide-spread usage of deep learning models has led to increasing number of minimally-interpretable models being used, as opposed to traditional models like decision trees, or even Bayesian and statistical machine learning models.

Explanations of models are also needed to reveal potential biases in the models themselves and assist with their debiasing.

This project aims to unpack the biases in models that may come from the underlying data, or biases in software (e.g. a simulation) that could be designed with a specific purpose and angle from the developers’ point-of-view. This project also aims to investigate techniques to generate diverse, robust and actionable explanations for a range of problems and data types and modality, from large-scale unstructured data, to highly varied sensor data and multimodal data. To this end, we look to generate counterfactual explanations that have a shared dependence on the data distribution and the local behaviour of the black-box model by level, and offer new metrics in order to measure the opportunity cost of choosing one counterfactual over another. We further aim to explore the intelligibility of different representations of explanations to diverse audiences through an online user study.

PUBLICATIONS

i-Align: An Interpretable Knowledge Graph Alignment Model, 2023

Salim, F., Scholer, F., et al.

Journal article

TransCP: A Transformer Pointer Network for Generic Entity Description Generation with Explicit Content-Planning, 2023

Salim, F., et al.

Journal article

Contrastive Learning-Based Imputation-Prediction Networks for In-hospital Mortality Risk Modeling Using EHRs, 2023

Salim, F., et al.

Conference paper

How Robust is your Fair Model? Exploring the Robustness of Diverse Fairness Strategies, 2023

Small, E., Chan, J., et al.

Journal article

Equalised Odds is not Equal Individual Odds: Post-processing for Group and Individual Fairness, 2023

Small, E., Sokol, K., et al.

Conference paper

Helpful, Misleading or Confusing: How Humans Perceive Fundamental Building Blocks of Artificial Intelligence Explanations, 2023

Small, E., Xuan, Y., et al.

Workshop paper

Navigating Explanatory Multiverse Through Counterfactual Path Geometry, 2023

Small, E., Xuan, Y., Sokol, K.

Workshop paper

Mind the gap! Bridging explainable artificial intelligence and human understanding with Luhmann’s Functional Theory of Communication, 2023

Sokol, K., et al.

Workshop paper

Measuring disentangled generative spatio-temporal representation, 2022

Chan, J., Salim, F., et al.

Conference paper

FAT Forensics: A Python toolbox for algorithmic fairness, accountability and transparency, 2022

Sokol, K., et al.

Journal article

Analysing Donors’ Behaviour in Non-profit Organisations for Disaster Resilience: The 2019–2020 Australian Bushfires Case Study, 2022

Chan, J., Sokol, K., et al.

Conference paper

BayCon: Model-agnostic Bayesian Counterfactual Generator, 2022

Sokol, K., et al.

Conference paper

RESEARCHERS

ADM+S Investigator Flora Salim

Prof Flora Salim

Lead Investigator,
UNSW

Learn more

Daniel Angus

Prof Daniel Angus

Chief Investigator,
QUT

Learn more

ADM+S Chief Investigator Paul Henman

Prof Paul Henman

Chief Investigator,
University of Queensland

Learn more

ADM+S Chief Investigator Mark Sanderson

Prof Mark Sanderson

Chief Investigator,
RMIT University

Learn more

Jeffrey Chan

Dr Jeffrey Chan

Associate Investigator,
RMIT University

Learn more

ADM+S Chief Investigator Falk Scholer

Prof Falk Scholer

Associate Investigator,
RMIT University

Learn more

ADM+S Investigator Damiano Spina

Dr Damiano Spina

Associate Investigator,
RMIT University

Learn more

ADM+S Investigator Maarten de Rijke

Prof Maarten de Rijke

Partner Investigator,
University of Amsterdam

Learn more

Peibo Li

Peibo Li

PhD Student,
UNSW

Learn more

Edward Small

Edward Small

PhD Student,
RMIT University

Learn more

Kacper Sokol

Kacper Sokol

Affiliate,
ETH Zurich

Learn more

PARTNERS

University of Amsterdam logo

University of Amsterdam

Visit website

Quantifying and Measuring Bias and Engagement

PROJECT SUMMARY

Man working on laptop

Quantifying and Measuring Bias and Engagement

Focus Areas: News & Media, Health
Research Programs: Machines, Data
Status: Active

Automated decision-making systems and machines – including search engines and intelligent assistants – are designed, evaluated, and optimised by defining frameworks that model the users who are going to interact with them. These models are typically a simplified representation of users (e.g., using the relevance of items delivered to the user as a surrogate for system quality) to operationalise the development process of such systems. A grand open challenge is to make these frameworks more complete, by including new aspects such as fairness, that are as important as the traditional definitions of quality, to inform the design, evaluation and optimisation of such systems.

Recent developments in machine learning, information access, and AI communities attempt to define mechanisms to minimise the creation and reinforcement of unintended cognitive biases.

However, there are a number of research questions related to quantifying and measuring bias and engagement that remain unexplored:
– Is it possible to measure bias by observing users interacting with search engines, or intelligent assistants?
– How do users perceive fairness, bias, or trust? How can these perceptions be measured effectively?
– To what extent can sensors in wearable devices and interaction logging (e.g., search queries, app swipes, notification dismissal, etc) inform the measurement of bias and engagement?
– Are the implicit signals captured from sensors and interaction logs correlated with explicit human ratings w.r.t. bias and engagement?

The research aims to address the research questions above by focusing on information access systems that involve automated decision-making components. By partnering with experts in fact-checking, we use misinformation management as the main scenario of study, given that bias and engagement play an important role in three main elements of the automated decision-making processes: the user, the system, and the information that is presented and consumed.

The methodologies considered to address these questions include lab user studies (e.g., observational studies), and the use of crowdsourcing platforms (e.g., Amazon Mechanical Turk). The data collection processes include: logging human-system interactions; sensor data collected using wearable devices; and questionnaires.

PUBLIC RESOURCES

Person working on laptop on wooden desk next to window

Open Source Software: Factchecking – Presentations

Target audience: Researchers, Software Developers
Code type: Python

View on Github

PUBLICATIONS

Human-AI Cooperation to Tackle Misinformation and Polarization, 2023

Spina, D., Sanderson, M., et al.

Journal article

Examining the Impact of Uncontrolled Variables on Physiological Signals in User Studies for Information Processing Activities, 2023

Ji, K., Spina, D., et al.

Conference paper

Can Generative LLMs Create Query Variants for Test Collections? 2023

Alaofi, M., Sanderson, M., et al.

Conference paper

Mitigating Negative Transfer with Task Awareness for Sexism, Hate Speech, and Toxic Language Detection, 2023

Spina, D., Rosso, P., Felipe Magnossão de Paula, A.

Conference paper

Do Social Media Users Change Their Beliefs to Reflect those Espoused by Other Users? 2023

Alknjr, H.

Conference paper

How do Human and Contextual Factors Affect the Way People Formulate Queries? 2023

Abu One, N.

Conference paper

Towards Detecting Tonic Information Processing Activities with Physiological Data, 2023

Ji, K., Hettiachchi, D., et al.

Conference paper

Ranking Interruptus: When Truncated Rankings Are Better and How to Measure That, 2022

Spina, D., et al.

Conference paper

Where Do Queries Come From? 2022

Alaofi, M., Spina, D., et al.

Conference paper

User-centered Non-factoid Answer Retrieval, 2022

Alaofi, M.

Conference paper

A Crowdsourcing Methodology to Measure Algorithmic Bias in Black-box Systems: A Case Study with COVID-related Searches, 2022

Scholar, F., Spina, D., Chia, H., Le, B.

Conference paper

AWARDS

2023 Pervasive and Ubiquitous Computing (UbiComp) International Symposium on Wearable Computing (ISWC)
Student Challenge Award
zzzGPT: An Interactive GPT Approach to Enhance Sleep Quality
Yonchanok (Pro) KhaokaewKaixin Ji, Marwah Alaofi, Hiruni Kegalle, Thuc Hanh Nguyen (UNSW) and Prof Flora Salim

2023 Pervasive and Ubiquitous Computing (UbiComp) International Symposium on Wearable Computing (ISWC)
Best Poster Award
Towards Detecting Tonic Information Processing Activities with Physiological Data’
Dr Damiano SpinaKaixin Ji, Prof Falk Scholer, Dr Danula Hettiachchi and Prof Flora Salim

17th Conference on Evaluation of Information Access Technologies (NTCIR-17)
Best Oral Presentation
Sachin Cherumanal Pathiyan

RESEARCHERS

Dr Damiano Spina

Dr Damiano Spina

Lead Investigator,
RMIT University

Learn more

ADM+S Chief Investigator Anthony McCosker

Assoc Prof Anthony McCosker

Chief Investigator,
Swinburne University

Learn more

ADM+S Investigator Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

ADM+S Chief Investigator Mark Sanderson

Prof Mark Sanderson

Chief Investigator,
RMIT University

Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University

Learn more

ADM+S Associate Investigator Jenny Kennedy

Assoc Prof Jenny Kennedy

Associate Investigator,
RMIT University

Learn more

ADM+S Chief Investigator Falk Scholer

Prof Falk Scholer

Associate Investigator,
RMIT University

Learn more

ADM+S Member

Nuha Abu Onq

PhD Student,
RMIT University

Marwah Alaofi

Marwah Alaofi

PhD Student,
RMIT University

Learn more

ADM+S Member

Hmdh Alknjr

PhD Student,
RMIT University

Sachin Pathiyan Cherumanal

PhD Student,
RMIT University

Learn more

Kaixin Ji

Kaixin Ji

PhD Student,
RMIT University

Learn more

PARTNERS

ABC logo

Australian Broadcasting Corporation

Visit website

AlgorithmWatch Logo

Algorithm Watch (Germany)

Visit website

Bendigo Health logo

Bendigo Hospital

Visit website

Google Logo

Google Australia

Visit website

RMIT ABC Fact Check Logo

RMIT ABC Fact Check

Visit website

Governing ADM Use

PROJECT SUMMARY

Blurred people in busy precinct

Governing ADM Use

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Program: Institutions
Status: Completed

The Governing ADM Use Project was an ‘umbrella’ project designed to seed the work of the ADM+S Institutions program in the rapidly evolving area of ADM and AI regulation. The project conceives the challenge of governing ADM use as a multi layered network incorporating the regulation of the use of ADM by government authorities, the regulation by government of ADM use in the commercial and private sector, and the interaction of ADM-specific regulation and governance with a range of other areas of law, regulation and governance that impinge and interact (more or less directly) with the specific governance of ADM/AI.

This latter category extends from data and privacy regulation to competition and consumer protection and beyond to sector and problem specific areas of regulation such as energy regulation, worker health and safety, labour force regulation and importantly environmental and planning laws. This program of work has sought to understand the special role of law as well as broader influences on public and private sector ADM use, and how these change – or need to change – to respond to the impacts of automation. A particular feature of this program of work has been to expand our understanding of the eco-system of law and governance properly concerned with regulating ADM/AI to include how we govern the ecological impact of ADM/AI use.

PUBLICATIONS

Harm to Nonhuman Animals from AI: a Systematic Account and Framework, 2023

Coghlan, S., Parker, C.

Journal article

Data problems and legal solutions – some thoughts beyond privacy, 2023

Weatherall, K., et al.

Book chapter

Reconstituting the Contemporary Corporation Through Ecologically Responsive Regulation, 2022

Parker, C., Haines, F.

Journal article

From ‘Corporate Governance’ to Ecological Regulation: Flipping the Regulatory Story on Climate Change, 2022

Parker, C.

Journal article

Algorithms as Figures. Towards a post-digital ethnography of algorithmic contexts, 2022

Cellard, L.

Journal article

The crimes of digital capitalism, 2022

Jiménez, A.

Journal article

Just Transitions in Australia: Moving Towards Low Carbon Lives Across Policy, Industry and Practice, 2022

Parker, C., Haines, F., et al.

Submission

More on Convening Technology: Blockchain, Fashion, and the Right to Know, 2022

Richardson, M., et al.

Journal article

Australian Competition and Consumer Commission, Digital Platform Services Inquiry Discussion Paper for Interim Report No 5: Updating competition and consumer law for digital platform services, 2022

Weatherall, K., et al.

Submission

Online Privacy Bill Consultation Submission, 2022

Goldenfein, J., Weatherall, K., Parker, C.

Submission

Submission in response to the Privacy Act Review, 2022

Weatherall, K., Trezise, M.

Submission

Submission to the Statutory Reviewer on the Consumer Data Right, 2022

Weatherall, K., Bednarz, Z., Dolman, C.

Submission

Submission on the Commonwealth Government Trusted Digital Identity Framework Position Paper, 2021

Weatherall, K.

Submission

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Lead Investigator,
University of Melbourne

Learn more

ADM+S Chief Investigator Megan Richardson

Prof Megan Richardson

Lead Investigator,
University of Melbourne

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
University of Melbourne

Learn more

Kim Weatherall

Prof Kim Weatherall

Chief Investigator,
University of Sydney

Learn more

ADM+S Investigator Karen Yeung

Prof Karen Yeung

Partner Investigator,
University of Birmingham

Learn more

José-Miguel Bello y Villarino

Dr José-Miguel Bello y Villarino

Research Fellow,
University of Sydney

Learn more

Henry Fraser

Dr Henry Fraser

Research Fellow,
QUT

Learn more

Loup Cellard

Dr Loup Cellard

Affiliate,
Datactivist Coop

Learn more

ADM+S Investigator Fiona Haines

Prof Fiona Haines

Affiliate,
University of Melbourne

Learn more

Aitor Jiménez

Dr Aitor Jiménez

Affiliate,
University of Melbourne

Learn more

Democratic Practices of Governance Given ADM

PROJECT SUMMARY

Busy street crossing aerial view

Democratic Practices of Governance Given ADM

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Program: Institutions
Status: Completed

This project examines possibilities for democratic practice, institutions and governance given automated decision-making (ADM). It focuses on challenges to and opportunities for liberal and democratic institutions and governance presented by ADM. The project aims to bridge analysis of ADM’s deployment across different domains with scholarly literature on republican and positive freedom, the rule of law and liberal democratic rights.

Overall, the project seeks to develop a theoretically rich analysis of democracy and freedom given ADM and apply the analysis to specific examples of current regulatory and democratic challenge.

PUBLICATIONS

Just Transitions in Australia: Moving Towards Low Carbon Lives Across Policy, Industry and Practice, 2022

Parker, C., Haines, F.

Submission

Privacy in the Republic, 2022

Kenyon, A.

Book

Countering hate speech in context: positive freedom of speech, 2022

de Silva, A., Kenyon, A.

Journal article

Law as Counterspeech, 2022

de Silva, A., et al.

Journal article

A Platformed Response to Hate Speech Against Women, 2022

de Silva, A.

Journal article

Introduction: Conceptualisations of Violence, 2022

de Silva, A., et al.

Book Chapter

The Crimes of Digital Capitalism, 2022

Jiménez, A.

Journal article

Law, Code and Exploitation: How Corporations Regulate the Working Conditions of the Digital Proletariat, 2022

Jiménez, A.

Journal article

The Australian News Media Bargaining Code, 2021

Goldenfein, J.

Analysis brief

Democracy of ExpressionPositive Free Speech and Law, 2021

Kenyon, A.

Book

Hate Speech Against Women: Addressing a Democratic Crisis, 2021

de Silva, A.

Policy brief

Positive Free Speech: A Democratic Freedom, 2021

Kenyon, A.

Book chapter

Surveillance Punitivism: Colonialism, Racism, and State Terrorism in Spain, 2021

Jiménez, A.

Journal article

Privacy, Punishment and Private Law, 2021

Roberts, A., Richardson, M.

Book chapter

Digital capitalism, what are the possible alternatives? 2021

Jiménez, A., et al.

Journal article

Adtech and children’s data rights, 2021

Archbold, L., Clifford, D., et al.

Journal article

Children’s Privacy in Lockdown: Intersections between Privacy, Participation and Protection Rights in a Pandemic, 2021

Archbold, L., Clifford, D., et al.

Journal article

Esports and the Platforming of Children’s During COVID-19, 2021

Fordyce, R., Archbold, L., et al.

Journal article

RESEARCHERS

ADM+S Investigator Christine Parker

Prof Christine Parker

Lead Investigator,
University of Melbourne

Learn more

Mark Andrejevic

Prof Mark Andrejevic

Chief Investigator,
Monash University

Learn more

Jake Goldenfein

Dr Jake Goldenfein

Chief Investigator,
Melbourne University

Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University

Learn more

ADM+S Investigator Sarah Erfani

Dr Sarah Erfani

Associate Investigator,
University of Melbourne

Learn more

Andrew Roberts

Prof Andrew Roberts

Associate Investigator,
University of Melbourne

Learn more

ADM+S Investigator Ivana Jurko

Ivana Jurko

Partner Investigator,
Red Cross Australia

Learn more

Anjalee de Silva

Dr Anjalee de Silva

Research Fellow,
University of Melbourne

Learn more

Chathurika Akurugoda

Chathurika Akurugoda

PhD Student,
University of Melbourne

Learn more

Lisa Archbold

Lisa Archbald

PhD Student,
University of Melbourne

Learn more

Phoebe Galbally

Phoebe Galbally

PhD Student,
University of Melbourne

Learn more

Avantik Tamta

Avantik Tamta

PhD Student,
University of Melbourne

Learn more

Aitor Jiménez

Dr Aitor Jiménez

Affiliate,
University of Melbourne

Learn more

ADM+S Investigator Fiona Haines

Prof Fiona Haines

Affiliate,
University of Melbourne

Learn more

ADM+S Chief Investigator Megan Richardson

Prof Megan Richardson

Affiliate,
University of Melbourne

Learn more

The Coronavirus Impact

PROJECT SUMMARY

COVID19 Stay safe on mobile device

The Coronavirus Impact

Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Program: Data
Status:
Completed

This project focused on a publication output: a themed issue of the journal New Media & Society. Our theme proposal was accepted and the theme issue is in its final stages pre-publication. We are still waiting for comments on one article, but 10 articles have been accepted for publication by the journal and the introduction has been written. We are only waiting for the decision on the last outstanding article before submitting the complete package to the editors for final review. The entire issue ended up being written by Centre members.

The focus of the issue is on the range of roles played by automated decision making systems in the pandemic response. These range from the automated curation of news content to automated contact tracing and air quality management. Contributions came from all four focus areas of the Centre. The timeframe for the issue enabled the inclusion of articles that tracked the shift from pandemic to endemic and an analysis of the ways in which systems developed in response to the pandemic persisted or faded away.

RESEARCHERS

Mark Andrejevic

Prof Mark Andrejevic

Lead Investigator,
Monash University

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Chief Investigator,
QUT

Learn more

ADM+S Chief Investigator Heather Horst

Prof Heather Horst

Chief Investigator,
Western Sydney University

Learn more

ADM+S Chief Investigator Sarah Pink

Prof Sarah Pink

Chief Investigator,
Monash University

Learn more

Gerard Goggin

Prof Gerard Goggin

Associate Investigator,
Western Sydney University

Learn more

Ariadna Matamoros Fernandez profile picture

Prof Ariadna Matamoros-Fernandez

Associate Investigator,
QUT

Learn more

Michael Richardson

Assoc Prof Michael Richardson

Associate Investigator,
UNSW

Learn more

Georgia Van Toorn

Dr Georgia van Toorn

Associate Investigator,
UNSW

Learn more

Rowan Wilken

Assoc Prof Rowan Wilken

Associate Investigator,
RMIT University

Learn more

Christopher O'Neill

Dr Christopher O’Neil

Research Fellow,
Monash University

Learn more

Dr Silvia Montaña-Niño profile picture

Dr Silvia Montaña-Niño

Affiliate,
University of Melbourne

Learn more

PARTNERS

Australian Red Cross Logo

Australian Red Cross

Visit website

OVIC Logo

Victorian Information Commissioner

Visit website

Data & Society Research Institute (US)

Visit website