AI Transparency

Twitter

Facebook

Copy Link

Key Facts

Key MethodsModel Cards, Datasheets

Required ByEU AI Act

Top News

Meta develops built-in AI Detector to identify synthetic content on its platforms

Meta is building an AI detection feature for its Meta AI chatbot that could identify synthetic content across its platforms. Discovered by TestingCatalog, the unreleased tool appears as an inactive menu option within Meta AI's interface. The development comes as the company faces criticism for contributing to AI slop while its Avocado models experience delays.

Technology3 sources

Meta develops built-in AI Detector to identify synthetic content on its platforms

Technology3 sources

Spotify hands control of AI recommendations to users with new Taste Profile feature

Spotify announced Taste Profile at SXSW, an AI-powered feature that lets Premium users personalize their algorithmic recommendations. Co-CEO Gustav Söderström revealed the beta will launch in New Zealand first, allowing listeners to review and customize how the app understands their music taste across genres, artists, and vibes.

Technology5 sources

Spotify hands control of AI recommendations to users with new Taste Profile feature

Technology5 sources

Peacock unveils AI features with Andy Cohen avatar, vertical NBA streams, and mobile games

NBCUniversal's Peacock is transforming its mobile app with AI-powered features launching this summer. The streaming platform introduces Your Bravoverse, an infinite scroll of Bravo clips narrated by an AI avatar of Andy Cohen, alongside vertical live NBA broadcasts and AI-enabled mobile games. With 44 million subscribers and room to grow, Peacock aims to compete with social media by turning passive viewing into interactive entertainment.

Technology7 sources

Peacock unveils AI features with Andy Cohen avatar, vertical NBA streams, and mobile games

Technology7 sources

Grammarly faces class-action lawsuit over AI Expert Review feature that mimicked writers without consent

Grammarly pulled its controversial AI Expert Review feature after journalist Julia Angwin filed a class-action lawsuit alleging the company used her identity and hundreds of other writers without permission. The tool, which simulated editorial feedback from figures like Stephen King and Carl Sagan, sparked backlash over unauthorized use of individuals' likenesses for commercial gain.

Policy7 sources

Grammarly faces class-action lawsuit over AI Expert Review feature that mimicked writers without consent

Policy7 sources

AI autocomplete covertly shifts human opinions on social issues, even when users ignore suggestions

Cornell University researchers discovered that AI autocomplete tools subtly manipulate how people think about major societal issues. In studies with over 2,500 participants, biased AI suggestions shifted opinions on topics like the death penalty and fracking—even among users who rejected the AI's text. Warning participants about AI bias beforehand or debriefing them afterward failed to prevent the attitude shift.

Science3 sources

AI autocomplete covertly shifts human opinions on social issues, even when users ignore suggestions

Science3 sources

Pentagon Investigates AI Role in Missile Strike That Killed 175 at Iranian School

The U.S. military is investigating whether AI played a role in a Tomahawk missile strike on an Iranian elementary school that killed at least 175 people, mostly children. Preliminary findings point to outdated targeting data, but questions persist about Anthropic's Claude AI, which the Pentagon uses for target selection despite designating the company a supply chain risk over its refusal to remove guardrails against autonomous weapons.

Policy2 sources

Pentagon Investigates AI Role in Missile Strike That Killed 175 at Iranian School

Policy2 sources

Pentagon reveals how Military AI chatbots accelerate targeting decisions in Iran operations

Defense officials confirm the US military is using AI chatbots like Claude to analyze intelligence and prioritize targets in Iran, striking 1,000 targets in the first 24 hours. The technology adds a conversational layer to Project Maven, but lawmakers are calling for stricter oversight as concerns grow about human judgment in war and the reliability of AI-powered decision support systems.

Technology7 sources

Pentagon reveals how Military AI chatbots accelerate targeting decisions in Iran operations

Technology7 sources

Anthropic launches research institute to study AI risks while battling Pentagon blacklist

Anthropic unveiled the Anthropic Institute, a new research body combining three existing teams to study how AI will reshape jobs, economies, and governance. The announcement comes as the company fights a Pentagon blacklist, triples its policy team, and opens its first Washington D.C. office this spring under new leadership.

Policy5 sources

Anthropic launches research institute to study AI risks while battling Pentagon blacklist

Policy5 sources

EU Parliament pushes for stricter AI copyright protections amid industry divide

The EU Parliament adopted recommendations for new EU rules to protect copyrighted work from AI training, proposing a European register listing all copyrighted materials used by AI models. The non-binding report calls for fair compensation for creators and transparency requirements, but tech lobby groups warn it could harm Europe's digital competitiveness.

Policy2 sources

EU Parliament pushes for stricter AI copyright protections amid industry divide

Policy2 sources

Oversight Board demands Meta overhaul AI-generated content rules after fake conflict video spreads

Meta's Oversight Board issued a sharp rebuke of the company's handling of AI-generated content, calling for a complete overhaul of its detection and labeling systems. The criticism follows a fake AI video depicting damage in Haifa during the Israel-Iran conflict that garnered over 700,000 views without proper labeling, exposing critical gaps in Meta's current approach to synthetic media moderation.

Policy6 sources

Oversight Board demands Meta overhaul AI-generated content rules after fake conflict video spreads

Policy6 sources

DOGE used ChatGPT to cancel $100 million in humanities grants under anti-DEI agenda

Elon Musk's Department of Government Efficiency deployed ChatGPT with a simple prompt to identify and terminate nearly 1,500 National Endowment for the Humanities grants. The AI-driven approach flagged projects ranging from Holocaust documentaries to Indigenous language archives as DEI-related, resulting in a clawback of over $100 million and triggering lawsuits alleging First Amendment violations.

Policy2 sources

DOGE used ChatGPT to cancel $100 million in humanities grants under anti-DEI agenda

Policy2 sources

Trump administration drafts strict AI guidelines as Pentagon designates Anthropic supply-chain risk

The Trump administration has drafted aggressive new AI guidelines requiring companies to permit any lawful use of their models by the U.S. government. The move follows the Pentagon's formal designation of Anthropic as a supply-chain risk, barring government contractors from using the AI firm's technology in military work after months of disputes over safeguards against surveillance and autonomous weapons.

Policy4 sources

Trump administration drafts strict AI guidelines as Pentagon designates Anthropic supply-chain risk

Policy4 sources

xAI loses legal challenge as California AI law forces data disclosure on training datasets

Elon Musk's xAI failed to block California's Assembly Bill 2013, which mandates AI companies disclose training data sources. U.S. District Judge Jesus Bernal ruled xAI hadn't proven the law violated free-speech rights or would reveal trade secrets, dealing a blow to the company's argument that disclosure could be economically devastating.

Policy4 sources

xAI loses legal challenge as California AI law forces data disclosure on training datasets

Policy4 sources

Mastercard launches Verifiable Intent to verify AI agent transactions with cryptographic proof

Mastercard introduced Verifiable Intent, an open-source framework designed to establish trust in agentic commerce. The system creates tamper-resistant records linking consumer identity, instructions, and transaction outcomes. Backed by Google, IBM, and Checkout.com, it integrates with Mastercard Agent Pay to provide cryptographic proof of authorization across AI-driven commerce platforms.

Technology2 sources

Mastercard launches Verifiable Intent to verify AI agent transactions with cryptographic proof

Technology2 sources

AI chatbots miss most medical diagnoses as millions seek health advice from ChatGPT

More than 40 million people consult ChatGPT daily for health information, but new research reveals AI chatbots correctly identify medical conditions only 34.5% of the time. Studies show these tools undertriage 52% of emergency cases and provide correct follow-up steps just 44.2% of the time, raising urgent questions about patient safety as AI healthcare becomes mainstream.

Health7 sources

AI chatbots miss most medical diagnoses as millions seek health advice from ChatGPT

Health7 sources

Prev 1…202122…38 Next