@article{coghlan_beyond_2026,

title = {Beyond ‘Basic’ AI-Animal Alignment},

author = {Simon Coghlan and Christine Parker},

url = {https://link.springer.com/10.1007/s13347-026-01039-y},

doi = {10.1007/s13347-026-01039-y},

issn = {2210-5433, 2210-5441},

year  = {2026},

date = {2026-03-01},

urldate = {2026-02-10},

journal = {Philosophy \& Technology},

volume = {39},

number = {1},

pages = {21},

abstract = {Abstract 

 In their article ‘AI Alignment: The Case for Including Animals’, Tse et al. compellingly argue for extending alignment endeavours beyond humans to also protect sentient animals. They call for AI alignment with a ‘basic’ level of animal welfare. Since ‘basic’ alignment carries minimal human cost, they argue, it can be widely accepted and is thus currently more strategically appropriate than is pursuing advanced or ideal alignment with animal welfare. This commentary paper argues that ‘basic’ AI alignment for animals is sometimes insufficient. It makes the case that going beyond basic AI-animal alignment is both feasible and morally required in a variety of circumstances.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{harms_developers_2026,

title = {Developers Are Central for Mitigation of AI Bias},

author = {Rebekah J. Harms and Rachel A. Ankeny and Lucy Carter and Aditi Mankad and Jackie Leach Scully},

url = {https://www.tandfonline.com/doi/full/10.1080/15265161.2025.2608629},

doi = {10.1080/15265161.2025.2608629},

issn = {1526-5161, 1536-0075},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-22},

journal = {The American Journal of Bioethics},

volume = {26},

number = {2},

pages = {112\textendash114},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{graham_internet_2026,

title = {On the Internet no-one knows you’re not a bot: ‘Botting’ on Reddit as participatory culture},

author = {Timothy Graham and Dominique Carlon},

url = {https://journals.sagepub.com/doi/10.1177/14614448251409210},

doi = {10.1177/14614448251409210},

issn = {1461-4448, 1461-7315},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-20},

journal = {New Media \& Society},

pages = {14614448251409210},

abstract = {Repetitive online communication is often labelled a ‘bot problem’ by platforms, policymakers and users. However, repetitive posting does not exclusively indicate automation; humans also engage in bot-like posting for various purposes. We adopt the term ‘botting’ to describe repetitive posting enacted through manual, semi-automated, or fully automated means. While emerging research has linked manual botting practices to commercial or fame-seeking motivations, we extend this scholarship by examining botting on Reddit \textendash a pseudonymous platform that lacks the affordances typically associated with monetisation or personal branding. Through a mixed-methods analysis, we examine a case study in which mass-scale, repetitive posting of the mushroom emoji emerged as ‘in-group’ behaviour within Reddit’s participatory culture, prompting a performative counterpublic response. Our findings challenge the binary between human and automated posting, and underscore the importance of situating research on AI-generated and automated content within the cultural and contextual frameworks that shape its production and reception.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{dehghan_entangled_2026,

title = {The entangled dynamics leading to the sedimentation of polarisation on political Reddit},

author = {Ehsan Dehghan and Dominique Carlon and Kateryna Kasianenko and Ashwin Nagappa and Vish Padinjaredath Suresh},

url = {https://www.tandfonline.com/doi/full/10.1080/1369118X.2026.2623523},

doi = {10.1080/1369118X.2026.2623523},

issn = {1369-118X, 1468-4462},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-20},

journal = {Information, Communication \& Society},

pages = {1\textendash24},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{sparrow_against_2026,

title = {Against Imaginary Friends: Why Digital Companions Are No Solution to Social Isolation},

author = {Robert Sparrow and James Brown},

url = {https://dl.acm.org/doi/10.1145/3750037},

doi = {10.1145/3750037},

issn = {0001-0782, 1557-7317},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-16},

journal = {Communications of the ACM},

volume = {69},

number = {2},

pages = {60\textendash68},

abstract = {For several decades, some people have suggested that loneliness and social isolation could be reduced by providing older persons with social robots. This project has, to date, failed: Social robots remain expensive, have limited functionality, and are unable to maintain the interest of users. However, a new technology, “digital companions,” looks more plausible. In this article, we extend the criticisms that have been made of social robots to digital companions and identify new dangers associated with the use of digital companions. It 

 is 

 possible that digital companions will allow people to feel less lonely. However, encouraging people to mistake imaginary friends for real friends is prima facie unethical. More importantly, encouraging people to have imaginary friends is no solution to social isolation. Proposals to use digital companions to combat loneliness and social isolation also underestimate the importance of touch, physical companionship, and mutual aid, when it comes to human well-being. 

 , 

 Relationships with "digital humans" might help people feel less lonely. But will this technology increase social isolation?},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{de_paula_mitigating_2026,

title = {Mitigating the Negative Transfer in Multi‐Task Learning for Harmful Language Detection in Spanish and Arabic},

author = {Angel Felipe Magnoss\~{a}o De Paula and Imene Bensalem and Damiano Spina and Paolo Rosso},

url = {https://onlinelibrary.wiley.com/doi/10.1111/exsy.70182},

doi = {10.1111/exsy.70182},

issn = {0266-4720, 1468-0394},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-20},

journal = {Expert Systems},

volume = {43},

number = {2},

pages = {e70182},

abstract = {ABSTRACT 

 Negative transfer continues to limit the benefits of multi‐task learning (MTL) in harmful language detection, where related tasks must share representations without diluting task‐specific nuances. We introduce task awareness (TA), a methodological framework that explicitly conditions MTL models on the task they must solve. TA is instantiated through two complementary mechanisms: Task‐aware input (TAI), which augments textual inputs with natural‐language task descriptions, and task embedding (TE), which learns task‐specific transformations guided by a task identification vector. Together they enable the encoder to disentangle shared and task‐dependent signals, reducing interference during joint optimisation. We integrate TA with BETO and AraBERT encoders and evaluate on six Spanish and Arabic datasets covering sexism, toxicity, offensive language, and hate speech. Across cross‐validation and official train‐test splits, TA consistently mitigates negative transfer, surpasses single‐task and conventional MTL baselines, and yields new state‐of‐the‐art scores on EXIST‐2021, HatEval‐2019, and HSArabic‐2023. The proposed methodology therefore combines a principled architectural innovation with demonstrated practical gains for multilingual harmful language detection. The resources to reproduce our experiments are publicly available at 

 https://github.com/AngelFelipeMP/Arabic‐MultiTask‐Learning 

 .},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{kasianenko_elements_2026,

title = {Elements of practices in digital publics: a model for identifying collective doings on social media platforms},

author = {Kateryna Kasianenko},

url = {https://academic.oup.com/ct/article/36/1/1/8307371},

doi = {10.1093/ct/qtaf026},

issn = {1050-3293, 1468-2885},

year  = {2026},

date = {2026-02-01},

urldate = {2026-03-03},

journal = {Communication Theory},

volume = {36},

number = {1},

pages = {1\textendash11},

abstract = {Abstract 

 Studies of public communication in digitally networked spaces have increasingly adopted the notion of practice to make sense of patterned doings in such spaces. They have identified and theorized about practices ranging from highly mundane and intimate doings to more collective and organized instances of grassroots platform governance to doings likely to result in harm. Despite this, it remains unclear how doings in networked publics must look, what properties they must have, who needs to engage in them, and for how long, to be considered practices. By synthesizing insights from practice theory and interdisciplinary Internet studies, I develop a model of practices in digital publics and outline interrelated elements of such sustained doings\textemdashaction, language, materiality, collectivity, knowledge, and normativity. This model can enable future empirical studies to define, identify, and differentiate practices more systematically within and across digitally networked spaces and consider their broader impact on digital citizenship.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{rousell_atmospheric_2026,

title = {Atmospheric Wellbeing: Sensing the More-Than-Human Dynamics of Air},

author = {David Rousell and Deborah Lupton},

url = {https://journals.sagepub.com/doi/10.1177/15327086251387108},

doi = {10.1177/15327086251387108},

issn = {1532-7086, 1552-356X},

year  = {2026},

date = {2026-02-01},

urldate = {2026-03-03},

journal = {Cultural Studies ↔ Critical Methodologies},

volume = {26},

number = {1},

pages = {3\textendash13},

abstract = {Clean air is vital to bodily, social, and planetary wellbeing. This article develops the concept of ‘atmospheric wellbeing’ as a framework for investigating the more-than-human dynamics of air through its affective and sensory qualities. Engaging the new field of critical air studies, the authors explore creative and multisensory social research methods which register the uneven distributions of air quality and the relationships between atmospheric sensing, feeling, and political action. This emerging approach offers new avenues for air quality research and seeds future-focused ideas for understanding and enhancing atmospheric wellbeing through creative means.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{kasianenko_community-led_2026,

title = {COMMUNITY-LED MODERATION IN ‘THE RUINS’ OF TWITTER/X: A CASE STUDY OF THE NORTH ATLANTIC FELLA ORGANIZATION (NAFO)},

author = {Kateryna Kasianenko and Adriana Matamoros-Fern\'{a}ndez and Olga Boichak},

url = {https://spir.aoir.org/ojs/index.php/spir/article/view/15188},

doi = {10.5210/spir.v2024i0.15188},

issn = {2162-3317, 2162-3317},

year  = {2026},

date = {2026-01-01},

urldate = {2026-03-25},

journal = {AoIR Selected Papers of Internet Research},

abstract = {Recent developments in platform governance point to community-led moderation as an increasingly preferred solution by mainstream social media platforms when dealing with problematic content. While research has shown the benefits of involving online communities in content moderation efforts, this type of moderation is also prone to limitations, such as leaving volunteer moderators on their own heuristics to decide upon complex categories of content. Using Anna Tsing's (2015) metaphor of 'life among ruins', in this paper we examine an online community called the North Atlantic Fella Organization (NAFO) that had to turn to volunteer governance in face of increasing failures of centralised content moderation on Twitter/X. Its members, also known as ‘fellas’, have been active on the platform since May 2022 by debunking and ridiculing online falsehoods spread by highly visible Russian government accounts and pro-Russian actors, reporting problematic behaviour, as well as fundraising on behalf of Ukraine. We identify three types of moderation practices relevant to the collective \textendash ‘soft-’, ‘hard-’, and ‘self’-moderation. While, for our participants, Russia’s war on Ukraine warranted such efforts, this community-led moderation required a high emotional and time investment on their behalf. Our findings attest to the ability of such practices to fill platform governance gaps, while also recognising the need for self-moderation and ongoing care for the community as vital for sustaining such practices.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{alaofi_use_2026,

title = {On the Use of LLMs for Relevance Labelling},

author = {Marwah Alaofi and Paul Thomas and Falk Scholer and Mark Sanderson},

url = {https://dl.acm.org/doi/10.1145/3788872},

doi = {10.1145/3788872},

issn = {1046-8188, 1558-2868},

year  = {2026},

date = {2026-01-01},

urldate = {2026-02-19},

journal = {ACM Transactions on Information Systems},

pages = {3788872},

abstract = {Large Language Models (LLMs) are increasingly used to replace human judges to assess the relevance of information objects, raising concerns about circularity, bias, and whether simulated preferences can substitute for human judgement. This work presents experiments using multiple LLMs to label passages for relevance. It examines their gullibility \textendash how easily they are misled into labelling irrelevant passages as relevant. It also compares LLMs with human judges in ranking systems, analysing differences in discriminative power and whether some systems benefit under LLM-based evaluation. Results show that LLMs are influenced by the presence of query terms, even with irrelevant or random passages. Moreover, LLM-generated rankings are highly correlated with those of human judges, with strong agreement on which system is better in pairwise comparisons. However, LLMs may exhibit lower discriminative power, as seen in flatter ranking slopes and missed significance for meaningful improvements. Yet, there are no cases where capable LLMs and human judges reach opposing conclusions with significance. LLMs may boost traditional systems more than neural ones, adding a new concern of system bias. These findings highlight the strong potential of LLMs for relevance labelling, while also highlighting failure cases that call for careful adoption and further research to maintain evaluation integrity.1 

 1},

keywords = {},

pubstate = {published},

tppubtype = {article}

}