AI Regulation

Twitter

Facebook

Copy Link

Key Facts

Key LawsEU AI Act

Key VoicesYoshua Bengio, Gary Marcus, Stuart Russell

Top News

Prediction markets bet Anthropic will restore Fable 5 access after government shutdown

Anthropic disabled its most powerful AI model Fable 5 on June 12 following a U.S. government order triggered by Amazon security findings. The model, which helped Stripe complete two months of coding work in one day, was pulled globally after researchers found vulnerabilities. Prediction market traders now give 74% odds that Fable 5 access returns by July 10.

Policy2 sources

Prediction markets bet Anthropic will restore Fable 5 access after government shutdown

Policy2 sources

Amazon challenges human-in-the-loop AI governance as tech giants shift to accountability models

Amazon's security VP Eric Brandwine argues that human-in-the-loop oversight fails because people stop paying attention—a phenomenon called normalization of deviance. Google, Microsoft, and IBM are also moving away from this model. Meanwhile, 82% of organizations discovered unauthorized AI agents in the past year, and 65% experienced AI-related security incidents, revealing that legacy infrastructure and excessive privileges create dangerous attack paths.

Technology16 sources

Amazon challenges human-in-the-loop AI governance as tech giants shift to accountability models

Technology16 sources

Anthropic Ban Triggers 30% Surge in Decentralized AI Tokens as Canada Warns of Centralization Risks

A US government order forced Anthropic to disable its Fable 5 and Mythos 5 models for all users worldwide, prompting Canadian Prime Minister Mark Carney to warn against overreliance on centralized AI providers. The shutdown triggered a rally in decentralized AI tokens, with Bittensor's TAO climbing 30% in 12 hours and the sector's market cap reaching $24.3 billion. Experts argue the incident exposes systemic risks in AI infrastructure.

Policy2 sources

Anthropic Ban Triggers 30% Surge in Decentralized AI Tokens as Canada Warns of Centralization Risks

Policy2 sources

Seattle Fire Department used AI to route 911 calls for two years without public knowledge

The Seattle Fire Department has been using Corti AI to listen to every 911 medical call since December 2023, prompting dispatchers to divert certain callers to a nurse line without public disclosure or surveillance ordinance review. The undisclosed deployment raises serious questions about transparency, civil liberties, and accountability in emergency services as AI spreads across Washington state.

Policy2 sources

Seattle Fire Department used AI to route 911 calls for two years without public knowledge

Policy2 sources

Meta tests military-grade facial recognition on smart glasses via Pentagon contractor partnership

Meta partnered with Rank One Computing, a Pentagon vendor serving US military and police, to test facial recognition technology on its Ray-Ban smart glasses. The dormant code, shipped to over 50 million phones, was deleted after public exposure. The partnership reveals how consumer tech merging with surveillance infrastructure blurs ethical boundaries in AI development.

Technology3 sources

Meta tests military-grade facial recognition on smart glasses via Pentagon contractor partnership

Technology3 sources

Family blames AI hospital system after woman dies waiting five days for ICU bed in Brazil

A 32-year-old Brazilian woman died after an AI-powered hospital management system allegedly delayed her ICU transfer for five days. Rebeca Cardoso Tenente Molina's family claims the Core-MG system assigned her a lower severity score than her deteriorating condition warranted, preventing doctors from exercising their medical judgment. The case highlights concerns about AI decision-making in critical healthcare situations.

Health2 sources

Family blames AI hospital system after woman dies waiting five days for ICU bed in Brazil

Health2 sources

Litigation emerges as real AI guardrail as lawsuits outpace regulation in shaping deployment

As AI models evolve faster than legislation can keep pace, litigation is emerging as the primary force shaping AI guardrails. The Mercor data breach, which resulted in seven class-action lawsuits, demonstrates how existing legal frameworks are being applied to AI issues without waiting for new AI-specific regulations. Experts argue that pragmatism driven by liability threats will curb harmful AI deployment far earlier than formal regulation.

Policy2 sources

Litigation emerges as real AI guardrail as lawsuits outpace regulation in shaping deployment

Policy2 sources

Apple built third-party AI system for Siri but kept it hidden at WWDC amid EU and legal pressure

Apple's iOS 27 developer beta reveals an Extensions framework allowing users to swap between ChatGPT, Claude, and Google Gemini inside Siri AI. The feature was built but never announced at WWDC, as Apple faces regulatory challenges with the EU Digital Markets Act, potential legal action from OpenAI, and concerns about undermining its own AI relaunch.

Technology2 sources

Apple built third-party AI system for Siri but kept it hidden at WWDC amid EU and legal pressure

Technology2 sources

Mark Carney warns US AI restrictions on Anthropic models expose risks of overreliance

Canadian Prime Minister Mark Carney said US restrictions on Anthropic's Fable 5 and Mythos 5 models demonstrate the dangers of depending on a limited number of American AI providers. Speaking ahead of the G7 summit, the former central banker drew parallels to the 2008 financial crisis, calling for diversification in AI infrastructure to avoid systemic vulnerabilities.

Policy5 sources

Mark Carney warns US AI restrictions on Anthropic models expose risks of overreliance

Policy5 sources

European Commission pushes back on US export controls blocking access to Anthropic AI models

The European Commission is scrutinizing US export controls that forced Anthropic to disable its most advanced AI models for foreign nationals, including Europeans. Brussels warns that national security measures should not discriminate against partners and emphasizes the need for Europe to strengthen its technological sovereignty in response to the abrupt access restrictions.

Policy4 sources

European Commission pushes back on US export controls blocking access to Anthropic AI models

Policy4 sources

States Push AI Regulation Forward Despite Trump's Executive Order Warning

Six months after President Trump issued an executive order warning states against regulating artificial intelligence, Republican and Democratic lawmakers are advancing state legislation anyway. With Congress stalled on federal oversight, states are targeting AI chatbots, employer use of AI systems, and catastrophic risks—filling a regulatory vacuum despite threats of legal challenges and funding restrictions.

Policy2 sources

States Push AI Regulation Forward Despite Trump's Executive Order Warning

Policy2 sources

Anthropic pulled powerful AI models offline after U.S. export control order over security fears

The Trump administration ordered Anthropic to restrict access to its Fable 5 and Mythos 5 AI models, citing national security concerns over their cybersecurity capabilities. The company responded by taking both models completely offline. Experts warn that similar AI capabilities will emerge from competitors regardless, and that export controls may hurt defenders more than attackers.

Policy50 sources

Anthropic pulled powerful AI models offline after U.S. export control order over security fears

Policy50 sources

Derbyshire Police officer faces criminal probe for allegedly using AI to fabricate evidence

A Derbyshire Police officer is under criminal investigation for allegedly using AI to create false evidence in multiple cases, marking the UK's first known incident of its kind. The officer has been removed from frontline duties while authorities work with the Crown Prosecution Service to assess potentially compromised cases. The investigation emerges just days after the UK launched PoliceAI, a national center backed by £75 million to guide responsible AI adoption in law enforcement.

Policy5 sources

Derbyshire Police officer faces criminal probe for allegedly using AI to fabricate evidence

Policy5 sources

Ukraine tested AI-powered drones that killed without human control in battlefield milestone

Ukrainian drone manufacturer reveals a 2024 test where fully autonomous drones with 'Terminator mode' killed Russian soldiers near Bakhmut without human intervention. The disclosure marks a watershed moment in warfare as AI-controlled drones made lethal decisions independently, while Ukraine simultaneously deploys AI-powered interceptor drones to defend against Russian Shahed drone attacks on cities.

Technology3 sources

Ukraine tested AI-powered drones that killed without human control in battlefield milestone

Technology3 sources

OpenText Ireland commits €105M AI investment to create 400 jobs in Cork and Galway

Canadian enterprise data management company OpenText announces €105 million investment in Ireland, creating 400 highly-skilled jobs over three years. The commitment doubles OpenText's presence in Ireland and marks the single largest investment by a Canadian tech company in Ireland, focusing on agentic AI, sovereign cloud solutions, and cybersecurity capabilities for EMEA markets.

Business2 sources

OpenText Ireland commits €105M AI investment to create 400 jobs in Cork and Galway

Business2 sources

Prev 1…192021…76 Next