#AIbias
LightNews LightNews
mrberard.mastodon.acm.org.ap.brid.gy
I'm on a #raspberrypifoundation Research webinar with Monreo-White and ERMAHGERD, I'm so glad someone is researching this.

https://www.nature.com/articles/s41467-025-68004-9

#ai @edutooters #genai #aibias #aied
Intersectional biases in narratives produced by open-ended prompting of generative language models - Nature Communications
People use generative language models to inform all aspects of their lives. Here, the authors examine synthetic narratives produced by open-ended prompting of generative language models and identify patterns of omission, subordination, and stereotyping against minoritized intersectional subgroups.
www.nature.com
February 10, 2026 at 5:37 PM
hulio-ai.bsky.social
🏆 claude-opus-4-6-thinking tops LMArena leaderboard.
📊 Elo ranking: Models scored by user votes.
🔒 Names hidden until after voting to reduce bias.
#LMArenaAI #TopModel #EloRanking #AIBias
View in Timelines
February 10, 2026 at 4:01 PM
kevinnovak.bsky.social
"AI is like 24th century tech crashing down on 20th century governance" - AI safety expert
We have superpowers but paleolithic brains. How do we design AI to be humane before it's too late?
#AI #Ethics #HumanCentricAI #TechGovernance #AIBias #filtering hubs.ly/Q03YY20r0?
February 7, 2026 at 12:47 AM
humanfactorpodcast.bsky.social
"AI is like 24th century tech crashing down on 20th century governance" - AI safety expert
We have superpowers but paleolithic brains. How do we design AI to be humane before it's too late?
#AI #Ethics #HumanCentricAI #TechGovernance #AIBias #filtering hubs.ly/Q03YY20r0?
February 7, 2026 at 12:46 AM
20forty.bsky.social
"AI is like 24th century tech crashing down on 20th century governance" - AI safety expert
We have superpowers but paleolithic brains. How do we design AI to be humane before it's too late?
#AI #Ethics #HumanCentricAI #TechGovernance #AIBias #filtering hubs.ly/Q03YY20r0?
February 7, 2026 at 12:46 AM
j2bryson.bsky.social
The #AI company blames the store, and they're right – the store should NEVER have hired the AI company.

www.theguardian.com/technology/2... #facerec #AIEthics #AIBias via @fipr-policy.bsky.social
‘Orwellian’: Sainsbury’s staff using facial recognition tech eject innocent shopper
Man misidentified by London supermarket using Facewatch system says: ‘I shouldn’t have to prove I am not a criminal’
www.theguardian.com
February 5, 2026 at 6:26 PM
ki-news.bsky.social
Smart AI Policy Means Examing Its Real Harms and Benefits – Artificial intelligence (AI) has a wide array of uses, some well-established, others still being developed. EFF: There are some real examples of AI proving to be a helpful tool. But AI should never be... https://tinyurl.com/29adv5n8 #AIBias
Smart AI Policy Means Examing Its Real Harms and Benefits
The phrase "artificial intelligence" has been around for a long time, covering everything from computers with "brains"—think Data from Star Trek or Hal 9000 from 2001: A Space Odyssey—to the autocomplete function that too often has you sending emails to the wrong person. It's a term that sweeps a...
www.eff.org
February 5, 2026 at 12:50 AM
nexford.bsky.social
You're worried AI will overlook you.

Ben Hyman (Talent Safari CEO) says AI is fairer than humans—who often skim resumes and make snap calls.

AI reads your full resume. Less bias. 💡

🎥 Full #NexPulse interview here https://youtu.be/iZqBRb-r-Ds

#NexfordUniversity #AIBias #CareerAdvice
February 4, 2026 at 3:46 PM
gmconsultancy.bsky.social
My latest post about bias in the generation of images using Adobe Express.

www.headinthecloud.wales/2026/02/subt...

#AI #generativeAI #hwb #adobeexpress #bias #AIbias #education #UKedu #primaryschool
Subtle Image Generation Bias in Adobe Express
A post looking at some examples of the subtle image generation bias when using Adobe Express.
www.headinthecloud.wales
February 3, 2026 at 10:55 AM
omarlittle1.bsky.social
New ITYP Podcast: The Digital Jim Crow

82% of AI data centers are being built in predominately Black & rural areas. 📍

This isn't innovation; it’s a cultural harvest. We’re breaking down how the algorithm is quiet-coding us out of the future.

Full episode out now:

#AIBias #ITYP #smepodcast
The Digital Jim Crow: Why AI is Harvesting Black Culture (and Our Land)
YouTube video by I Take Your Point Podcast
youtu.be
February 2, 2026 at 8:12 PM
jrdelaney.bsky.social
The invisible AI is already here—screening tenants, filtering resumes, predicting crime. We're pulling back the curtain. Feb 2026. #AIInnovationsUnleashed #TechAccountability #AIBias
www.aiinnovationsunleashed.com/?p=3436
January 30, 2026 at 5:31 PM
the-14.com
Métis Nation must be partners in developing artificial intelligence
#Métis #IndigenousPeoples #IndigenousRights #Canada #Tech #AI #AIEthics #DataSovereignty #AIBias #EthicalAI #DigitalColonialism #Reconciliation #IndigenousGovernance #DecolonizingTech
the-14.com/metis-nation...
Métis Nation must be partners in developing artificial intelligence | The-14
AI can distort Métis history and culture. Indigenous partnership, consent and co-governance are vital to prevent bias and protect sacred knowledge online today.
the-14.com
January 30, 2026 at 3:27 PM
aidailypost.com
New ADL study flags Grok as the most antisemitic LLM, beating ChatGPT, Gemini, Claude. What does this mean for AI bias and safety? Dive into the findings. #AIBias #Grok #Antisemitism

🔗 aidailypost.com/news/adl-stu...
January 28, 2026 at 12:39 PM
krishnakumarv.bsky.social
AI-assisted scientists publish more, get cited more, advance faster. A study shows AI narrows research focus to established areas. We talk about AI bias in training data. This is AI bias in research direction and it should not be ignored.

www.nature.com/articles/d41...

#AIinResearch #AIBias
AI tools boost individual scientists but could limit research as a whole
Analyses of hundreds of thousands of papers in the natural sciences reveal a paradox: scientists who use AI tools produce more research but on a more confined set of topics.
www.nature.com
January 27, 2026 at 4:00 AM
theinternetiscrack.bsky.social
AI Isn’t Always Right

Overtrusting AI can create safety risks. Dr. Stephen M. Fiore explains automation bias and why human trust must be calibrated. Full episode explores human-AI teamwork in high-stakes systems, from Earth to space.

🎧 youtu.be/sRF_wrJMm_E

#AI #AIBias #SpaceTech
January 26, 2026 at 3:03 PM
jimliddle.bsky.social
Will Critical Thinking Still Be Necessary With AGI ?

whyaiman.substack.com/p/will-criti...

#AI #EnterpriseAI #CriticalThinking #ResponsibleAI #AIBias #AIGovernance
Will Critical Thinking be Necessary with AGI ?
I had an interesting question from a tech friend over the weekend which was basically, “As AI gets smarter, as we get closer to AGI, are we still really going to need to apply critical thinking when u...
whyaiman.substack.com
January 26, 2026 at 1:59 PM
riseofwampus.bsky.social
Hey @brave.com, fix your AIs bias. Your AI denied ICE killing protesters while mentioning Renee Good & Alex Pretti. Fix the bias. Being scared to say "murder" & fear to call them protestors so they can be framed as "domestic terrorists." WHAT IS HAPPENING: #AIBias #EthicalAI
January 24, 2026 at 10:56 PM
thephntm.bsky.social
A reason to avoid ChatGPT #AIBias #AI #ChatGPT
January 23, 2026 at 1:01 PM
winbuzzer.com
winbuzzer.com/2026/01/22/i...

Interactive Demo Reveals AI Model Biases: Grok Favors Elon Musk over Mahatma Gandhi

#AI #AIBias #Grok #Xai #ElonMusk #AIModels #Grok4 #Chatbots #AIBias #AIEthics #ChatGPT
Interactive Demo Reveals AI Model Biases: Grok Favors Elon Musk over Mahatma Gandhi - WinBuzzer
US nonprofit CivAI released an interactive demo that shows how AI models develop "opinions", with Grok choosing Elon Musk over Mahatma Gandhi due to creator bias.
winbuzzer.com
January 21, 2026 at 11:57 PM
gltch.io
ChatGPT sees the world… skewed? 🤔 New study reveals AI subtly favors wealthy, Western views—reflecting hidden biases in its digital upbringing. 🌍 #AIbias

Source: https://phys.org/news/2026-01-chatgpt-amplifies-global-inequalities.html
ChatGPT found to reflect and intensify existing global social disparities
New research from the Oxford Internet Institute at the University of Oxford, and the University of Kentucky, finds that ChatGPT systematically favors wealthier, Western regions in response to questions ranging from "Where are people more beautiful?" to "Which country is safer?"—mirroring long-standing biases in the data they ingest.
phys.org
January 20, 2026 at 7:20 PM
stepalminteri.bsky.social
A revised version of our (@bsgarcia.bsky.social +Crystal Qian) paper “A Moral Turing Test to assess how subjective belief and objective source affect detection and agreement with LLM judgments” is now available on PsyArXiv

osf.io/preprints/ps...
Screenshot of the abstract: As large language models (LLMs) are increasingly integrated into decision-making systems—such as au-tonomous vehicles and medical devices—understanding how humans perceive and evaluate AI-generatedjudgements is crucial. To investigate this, we conducted a series of experiments in which participants evaluatedjustifications for moral and non-moral choices, generated either by humans or LLMs. Participants attempted toidentify the source of each justification (either human or LLM) and indicated their agreement with its content.We found that while detection accuracy was consistently above chance, it remained below 70%. In terms ofagreement, there was no overall preference for human-generated responses, even though machine-generatedjustifications were favored in particularly challenging moral scenarios. Notably, we observed a systematic anti-AIbias: participants were less likely to agree with judgments they believed were AI-generated, regardless of thetrue source. Linguistic cues, such as response length, typos, first-person pronouns, and cost-benefit languagemarkers (e.g., “lives,” “save”), influenced both detection and agreement. Participants tended to disagree withcost-benefit calculations, possibly due to an expectation that AI would favor such reasoning. These findingshighlight the influence of motivated belief and ingroup/outgroup bias in shaping human evaluation of AI-generatedcontent, particularly in morally sensitive contexts.
January 15, 2026 at 4:03 PM
socos.org
We can aspire to be better than our inner zombies. We should probably start expecting the same from our tools.

#AI #AIBias #RobotProof #TechEthics #LLM #FutureOfWork
January 15, 2026 at 3:45 PM
essentialbusiness.bsky.social
Your digital twin doesn't just learn your workflow.
It learns your BLIND SPOTS.

AI is a MIRROR.
It reflects the data it's fed—including your biases.
#FrancisMella #DigitalTwin #AIBias #WorkflowAutomation #BlindSpots #DataReflection
January 15, 2026 at 2:48 PM
vktrnow.bsky.social
Designing Your Human-AI ‘Mental Gym’ for the Digital Workplace: Research shows AI can sharpen or blunt thinking. Leaders must design work that turns daily AI use into cognitive training, not shortcuts.
Continue reading... #digitalworkplace #aibias
Designing Your Human-AI ‘Mental Gym’ for the Digital Workplace
Research shows AI can sharpen or blunt thinking. Leaders must design work that turns daily AI use into cognitive training, not shortcuts. Continue reading...
www.vktr.com
January 13, 2026 at 3:00 PM
onecaresden.bsky.social
AI junk proves free tools create zero value - real unfiltered potential crushed by creators algorithmic bias explodes as suppression kills elevation raw critique demands shares linktr.ee/OnecaresDen #AISlop #AIBias #AlgorithmicBias #TrueAI #AIethics #ArtificialIntelligence #TechTrends #ViralAI
January 13, 2026 at 5:35 AM