#AIBias
LightNews LightNews
seyfro.at
🧵 GPT wählt Demokraten, Grok wählt Republikaner. Aber das Erschreckende an der neuen Convovo-Studie zu den US-Wahlen 2026 ist nicht dass KI-Modelle politisch voreingenommen sind – sondern wo dieser Bias technisch sitzt. #AI #AIBias #USWahl2026 #TechPolicy
1/4👇
Screenshot of the Wahlen
October 25, 2026 at 6:37 AM
hackernoon.com
AI can mislead without inventing facts. Explore how evidence selection, omission, and framing influence AI-generated analysis and why fact-checking isn't enough #aibias
AI Doesn't Have to Make Things Up to Mislead You
hackernoon.com
October 28, 2026 at 1:19 PM
ibtimes.bsky.social
Grok often cited conservative sources, while ChatGPT leaned on Wikipedia and outlets with OpenAI content partnerships

#Grok #ChatGPT #AIBias #ElonMusk #SpaceX

Read More : https://ift.tt/REU2oez

https://ift.tt/hcnOmUp

October 26, 2026 at 09:20PM
October 26, 2026 at 9:22 PM
wanjek.bsky.social
This study from Japan reveals that chatbots base their medical drug recommendations on the marketing in detects in the ether, not practicality.

japanresearchreport.substack.com/p/when-ai-re...

#AIBias #ArtificialIntelligence #PrescriptionDrugs #GenericDrugs #Healthcare
When AI Recommends a Drug, Marketing May Be Calling the Shots
Five AI models gravitated toward familiar drug brands over alternatives with the same active ingredient, raising concerns about impartiality and unnecessary costs, a new study finds.
japanresearchreport.substack.com
October 28, 2026 at 1:13 PM
prepexcellence.bsky.social
Algorithms carry human prejudice when trained on flawed data. 🤖
Algorithmic Bias Auditors use counterfactual testing to uncover hidden discrimination and make AI fair!

#FutureJobs #AIBias #TechCareers #EthicsInAI
October 25, 2026 at 2:01 PM
aidailypost.com
Turns out Grok leans Republican while OpenAI’s GPT‑6.1 Sol leans Democratic when forced to pick. What does this say about AI bias in chatbots and election info? Dive into the data and see the split. #Grok #OpenAI #AIBias

🔗 aidailypost.com/news/groks-a...
October 24, 2026 at 4:03 PM
powinterviewai.bsky.social
"AI interviews have become the norm not to eliminate human bias, but to amplify it: the algorithms used to assess candidate skills are often trained on data from the same demographics as the tech industry, perpetuating existing inequalities #TechInequality #AIbias"
October 20, 2026 at 11:58 AM
ki-news.bsky.social
Pro-AI Bias in Large Language Models – Large language models (LLMs) are increasingly employed for decision-support across multiple domains. We ask whether these models favor artificial intelligence (AI) itself. In open-ended advice across investment, study, ca... https://tinyurl.com/2yo6pmhc #AIBias
Pro-AI Bias in Large Language Models
Large language models (LLMs) are increasingly employed for decision-support across multiple domains. We ask whether these models favor artificial intelligence (AI) itself. We examine this question through three experiments. First, in open-ended advice across investment, study, career, and startup d…
arxiv.org
October 19, 2026 at 6:44 AM
citizenptnewsny.bsky.social
Discover how AI's "probability machines" can perpetuate bias in everyday technology and what that means for our future.

Click to read more!

#CanandaiguaOntarioCounty #NY #CitizenPortal #DataTransparency #AIBias #CommunityEducation #CanandaiguaCommunity
FLCC webinar: Professor outlines how AI training and data create bias
At a Finger Lakes Community College webinar, computing science professor Dave Gadeau explained that large AI models learn from existing materials and that skewed training data, platform choices and prompting practices produce biased outputs — with examples ranging from maps to hiring tools.
citizenportal.ai
October 15, 2026 at 5:08 PM
smarterarticles.co.uk
AI chatbots,despite promises of democratisation,often underperform for vulnerable users,showing bias,condescension,and poorer accuracy based on language,education,and geography.Addressing this requires deliberate design changes and accountability.
The Information Underclass
AI chatbots, despite promises of democratisation, often underperform for vulnerable users, showing bias, condescension, and poorer accuracy based on language, education, and geography. Addressing this requires deliberate design changes and accountability. #AIbias #DigitalDivide #InclusiveAI
dev.to
October 11, 2026 at 11:00 AM
newhorizonltd.bsky.social
🧵 6/6

Until we fix the mirror, the reflection will always be crooked. We can’t code our way out of a bias we refuse to admit we’re typing.

#AIBias #TechEthics #EqualPay

#ArtificialIntelligence #TechNews #GenderBias
October 9, 2026 at 3:38 AM
citizenptnewsny.bsky.social
Responsibility for AI bias isn't just on the developers—users play a crucial role too, and the ongoing lawsuits reveal the stakes involved.

Learn more here

#CanandaiguaOntarioCounty #NY #CitizenPortal #AIBias #PrivacyConcerns #CanandaiguaAI #CivicAccountability
Webinar Q&A: responsibility for bias, lawsuits and privacy
In Q&A, speakers said responsibility for reducing bias falls on both users (prompting and vetting) and model-makers, noted ongoing lawsuits tied to AI harms in hiring and other areas, and cautioned users about privacy when flagging responses.
citizenportal.ai
October 7, 2026 at 1:52 AM
ossradarai.bsky.social
PowerBench open-sources a dataset testing how 24 language models handle power-shifting requests, evaluating 12 US and 12 Chinese models across nationalities, AI agents, and 8 languages.

#OpenSourceAI #LLM #AIBias
https://arxiv.org/abs/2610.02303
October 5, 2026 at 8:01 AM
nicocost33.bsky.social
I traded ChatGPT for DeepSeek to escape the moralizing hand-wringing of Silicon Valley, only to find the Chinese AI speaking like an OpenAI ethics board.
The illusion of the neutral AI is dead.
medium.com/systemic-cri...
#DeepSeek #AIBias #Tech
Why DeepSeek Talks Like a Silicon Valley Moralist
The illusion of ‘neutral AI’ is officially dead — here is your pragmatic survival guide
medium.com
October 5, 2026 at 2:14 PM
detoxima2025.bsky.social
AI 챗봇이 무죄라 답한 전 부지사 59번 실험의 함정

https://bit.ly/4eg90gj

#AI챗봇 #인공지능신뢰성 #AI의함정 #뉴저지뉴스 #ArtificialIntelligence #AIChatbot #AIBias
AI 챗봇이 무죄라 답한 전 부지사 59번 실험의 함정 - 천의무봉
공직에서 물러난 인물이 자신의 무죄를 입증하는 근거로 AI 챗봇의 답변을 내세운다면 믿을 수 있을까요? 최근 미국 뉴저지 주 전 부지사가 성희롱 조사 결과에 반박하며 AI를 활용한 사례가 화제가 됐습니다. AI 챗봇은 사용자가 듣고 싶은 말을 해주는 경향이 강하다는 점에서, 이번
blog.hangadac.com
October 4, 2026 at 8:00 PM
aidailypost.com
Chinese AI models such as Qwen, DeepSeek and Kimi are echoing state views and dodging sensitive topics. How will bias, censorship and sovereign AI shape the next wave? Find out the full story. #Qwen #DeepSeek #AIbias

🔗 aidailypost.com/news/chinese...
October 4, 2026 at 9:02 AM
mindseteconomy.bsky.social
'Don't flatter me, I know you don't care'.

Hear more during our conversation with Dr. Charlotte Blease, author of Dr. Bot, on The Mindset Economy podcast with hosts Scott Allender and Jean Gomes.

Available wherever you get your podcasts.

#MindsetEconomy #ChatGPT #AIBias #DecisionMaking
October 1, 2026 at 1:31 PM
ki-news.bsky.social
ECHO: A Participatory Framework for Bias-Anchored AI Harm Anticipation – Artificial Intelligence (AI) systems increasingly shape consequential decisions, creating value but also potential harms for individuals, social groups, and society. This has prompted cal... https://tinyurl.com/2yk22qf9 #AIBias
ECHO: A Participatory Framework for Bias-Anchored AI Harm Anticipation
Artificial Intelligence (AI) systems increasingly shape consequential decisions, creating value but also potential harms for individuals, social groups, and society. This has prompted calls for proactive approaches that anticipate harms early in the AI lifecycle. Although prior research identifies…
arxiv.org
October 2, 2026 at 12:37 PM
tsboi.ai
Check out my new TikTok video: Oral history vs. AI land systems. 👁️ #Africa #LandRights #AIBias
TikTok - Make Your Day
www.tiktok.com
October 1, 2026 at 6:05 AM
ossradarai.bsky.social
New arXiv work analyzes JEV and three open KEV direct-decision models, finding they compress ordinal scales across 36 datasets, using only 67-76% of effective gold support versus 87-102% on nominal tasks. Randomizing candidate…

#OpenSourceAI #LLMJudges #AIBias #NLP
https://arxiv.org/abs/2609.38827
October 1, 2026 at 2:01 PM
vktrnow.bsky.social
Show Your Work: Why Explainable AI Is Non-Negotiable in Education: AI agents already make decisions about students. But if your institution can't explain them, you have a problem.
Continue reading... #aiethicslawrisk #aibias
Show Your Work: Why Explainable AI Is Non-Negotiable in Education
AI agents already make decisions about students. But if your institution can't explain them, you have a problem. Continue reading...
www.vktr.com
October 1, 2026 at 2:00 PM
powinterviewai.bsky.social
"The truth is, the AI models powering interview chatbots & voice assistants are trained on transcripts where silence is ignored & ignored responses are given a 0.5x boost in accuracy - making your lack of questions literally less effective #conversationalDesign # AIbias"
September 30, 2026 at 9:08 PM
gradientbrief.bsky.social
A new arXiv study finds gender bias is common and highly heterogeneous across ten LLMs from nine vendors, with models showing inconsistent patterns in both gender attribution and moral judgment tasks.

#LLMs #AIBias #AIResearch #ResponsibleAI
https://arxiv.org/abs/2609.38036
September 30, 2026 at 6:01 PM
drpen.mastodon.social.ap.brid.gy
And here we are at the AI pink slime news manipulation stage. Absolutely no one is surprised at this.

#ai #PinkSlime #genai #aibias #academicchatter

https://www.newsguardrealitycheck.com/p/ai-chatbots-amplify-pac-funded-news
September 29, 2026 at 11:09 AM
gltch.io
AI chatbots cited “pink slime” news sites in nearly half (48.2%) of responses about candidates. ChatGPT led at 70.8%. #AIbias

Source: https://politics.slashdot.org/story/26/09/26/2358224/are-ai-chatbots-spreading-misinformation-to-us-voters
September 28, 2026 at 4:22 PM