Responsible AI

Twitter

Facebook

Copy Link

Key Facts

Key BodiesNIST, ISO, OECD, UNESCO

Key FrameworksNIST AI RMF, ISO 42001, EU AI Act

Top News

New York Times faces union backlash over AI tools monitoring employee performance

Unionized staff at The New York Times have filed grievances and unfair labor practices charges, alleging the company uses AI tools to monitor and evaluate employee performance without proper disclosure. The Times Tech Guild claims management violated their collective bargaining agreement by deploying DX and Glean, two internal systems that track work output and access internal documentation, raising concerns about surveillance and arbitrary performance metrics.

Policy2 sources

New York Times faces union backlash over AI tools monitoring employee performance

Policy2 sources

Shadow AI drives security incidents at 58% of firms as workers bypass approved tools

More than half of organizations experienced AI-related security incidents in the past year, driven by shadow AI use. New research reveals 52% of knowledge workers use unapproved AI tools, while 64.5% of personal AI account activity is actually for work purposes. Despite this, 90% of executives remain confident in their visibility over AI usage, exposing a dangerous disconnect between leadership perception and workplace reality.

Technology2 sources

Shadow AI drives security incidents at 58% of firms as workers bypass approved tools

Technology2 sources

Cisco finds no closed frontier AI model safe from multi-turn attacks across major providers

Cisco's AI Threat Research team tested 15 proprietary flagship models from OpenAI, Anthropic, Google, Amazon, and xAI, revealing multi-turn attack success rates between 7.9% and 88.3%. The study exposes how single-turn safety benchmarks fail to capture real-world adversarial behavior, where attackers iterate and adapt across conversations. Even the strongest models like Claude showed vulnerability, with rates climbing from 2.2% to 16.2% under iterative pressure.

Technology2 sources

Cisco finds no closed frontier AI model safe from multi-turn attacks across major providers

Technology2 sources

AI models favor Catholicism and sideline faith when users need guidance on grief and moral decisions

A multi-university consortium found that Large Language Models systematically exclude religious perspectives from answers about grief, forgiveness, and ethics—mentioning faith only 5-16% of the time when Americans expect it 45-59% of the time. Every AI model tested showed conversion bias, steering users toward Catholicism while displaying negative bias against Jehovah's Witnesses. The findings emerged as Pope Leo XIV released an encyclical warning that AI absorbs the values and blind spots of its creators.

Technology5 sources

AI models favor Catholicism and sideline faith when users need guidance on grief and moral decisions

Technology5 sources

82% of Teachers Navigate AI in Education Without Formal Guidance as Schools Struggle to Keep Pace

A new Gallup and Walton Family Foundation study reveals that 82% of K-12 teachers receive no formal guidance on AI use despite six in 10 using AI tools for work. The absence of clear policies leaves educators navigating AI in schools alone, contributing to increased teacher burnout and job dissatisfaction as the expectations gap widens.

Policy2 sources

82% of Teachers Navigate AI in Education Without Formal Guidance as Schools Struggle to Keep Pace

Policy2 sources

Ozzy Osbourne AI avatar announced by family, sparking debate over recreating deceased celebrities

Sharon and Jack Osbourne unveiled plans for an AI-powered hologram of the late heavy metal icon at Licensing Expo 2026. The digital recreation will use Ozzy's voice and image to interact with fans in real-time through Proto Luma units launching this summer. While the family insists the project honors his legacy, critics question the ethics of commercializing deceased celebrities through AI technology.

Entertainment3 sources

Ozzy Osbourne AI avatar announced by family, sparking debate over recreating deceased celebrities

Entertainment3 sources

Gartner warns 40% of enterprises will decommission AI agents by 2027 due to governance failures

Gartner predicts two in five companies will have to scale back their autonomous AI agents by 2027 because of critical gaps in governance frameworks. The root cause? Organizations treat AI governance as binary—either fully locked down or completely trusted. This uniform approach creates security vulnerabilities and operational risks that often surface only after production incidents occur.

Technology3 sources

Gartner warns 40% of enterprises will decommission AI agents by 2027 due to governance failures

Technology3 sources

Startups tap India's gig workers to collect physical AI training data, igniting privacy concerns

Silicon Valley startup Human Archive has raised $8.2 million to deploy camera-equipped headsets on Indian gig workers, collecting first-person video of everyday tasks to train robots. The initiative has sparked controversy as home services platforms like Pronto pilot in-home data recording, prompting India's IT ministry to take notice while legal experts warn existing privacy laws may not adequately address AI systems learning inside private spaces.

Technology7 sources

Startups tap India's gig workers to collect physical AI training data, igniting privacy concerns

Technology7 sources

Pope Leo XIV releases major AI encyclical calling for 'disarmament' of artificial intelligence

Pope Leo XIV unveiled Magnifica Humanitas, his first encyclical addressing artificial intelligence. The 40,000-word document calls for AI to be 'disarmed' from monopolistic control and warns against autonomous weapons, neo-colonial data collection, and concentration of power. Presented alongside Anthropic co-founder Chris Olah, the encyclical updates Catholic social teaching for the AI age.

Policy87 sources

Pope Leo XIV releases major AI encyclical calling for 'disarmament' of artificial intelligence

Policy87 sources

AI Hallucinations Infiltrate Research Papers as 146,000 Fake References Enter Scientific Record

A crisis in academic integrity is unfolding as AI hallucinations contaminate scientific literature at an alarming rate. Over 146,000 fabricated references generated by AI tools entered research papers in 2025 alone, with the vast majority bypassing peer review. The surge has sparked heated debate after arXiv announced bans for authors who fail to verify AI-generated citations, prompting backlash from academics who argue the policy is too strict.

Policy3 sources

AI Hallucinations Infiltrate Research Papers as 146,000 Fake References Enter Scientific Record

Policy3 sources

Barnes & Noble CEO's stance on AI-generated books triggers backlash from authors and readers

Barnes & Noble CEO James Daunt sparked controversy by stating the bookseller won't ban AI-written books as long as they're labeled. The comments drew swift backlash from authors and readers who view AI technology as built on stolen writing. Daunt later clarified that Barnes & Noble actively excludes AI-generated books from its catalogue, calling it a 'straightforward rejection.'

Entertainment3 sources

Barnes & Noble CEO's stance on AI-generated books triggers backlash from authors and readers

Entertainment3 sources

Anthropic plans public release of Mythos AI after discovering 10,000+ security vulnerabilities

Anthropic revealed plans to release its restricted Mythos AI model to the public once adequate safeguards are developed. Through Project Glasswing, the model has already uncovered more than 10,000 high-severity flaws across critical software infrastructure in just one month, overwhelming security teams with the sheer volume of bugs requiring patches.

Technology12 sources

Anthropic plans public release of Mythos AI after discovering 10,000+ security vulnerabilities

Technology12 sources

Author caught using AI in writing says he can't quit the technology despite fabricated quotes

Steven Rosenbaum's book about AI and truth was found to contain over half a dozen fabricated or misattributed quotes created by AI tools. Despite admitting the errors and facing widespread criticism, the author insists he will continue using AI in his writing process, calling it "magical" and a "delightful writing companion" that he cannot abandon.

Entertainment6 sources

Author caught using AI in writing says he can't quit the technology despite fabricated quotes

Entertainment6 sources

Sadiq Khan Blocks £50m Palantir Met Police Contract Amid Procurement and Ethics Concerns

London Mayor Sadiq Khan blocked a £50m contract between the Metropolitan Police and US tech firm Palantir, citing procurement rule breaches and value-for-money concerns. The decision has sparked fierce debate within Labour, with Palantir's UK chief accusing Khan of prioritizing politics over public safety while MPs defend the move as protecting London's values.

Policy3 sources

Sadiq Khan Blocks £50m Palantir Met Police Contract Amid Procurement and Ethics Concerns

Policy3 sources

Universal Music Group and TikTok renew deal to combat unauthorized AI music on platform

Universal Music Group and TikTok have signed a multi-year strategic licensing agreement that commits both companies to removing unauthorized AI-generated music from the platform while improving artist attribution. The deal marks a significant shift after UMG pulled its catalog in 2024 over royalty disputes and AI concerns, now establishing a framework for how platforms handle AI music and intellectual property rights.

Business3 sources

Universal Music Group and TikTok renew deal to combat unauthorized AI music on platform

Business3 sources

Prev 1…242526…86 Next