China unveils draft export controls covering AI models, chip designs, biotech and solar tech
Beijing's proposed list would restrict outbound transfers of advanced AI and related technologies amid escalating US-China trade tensions.
The Next Web · +8 outlets
Eight US states sue OpenAI, alleging ChatGPT harms children and deceives adults
The coordinated state lawsuits add to mounting legal pressure on OpenAI over chatbot safety and consumer protection.
WLFI · +10 outlets
Google opens Gemini 4 Argon, scoring 97.5 percent on ARC-AGI-3
The result makes Google the second lab after OpenAI to cross the 95 percent mark on the frontier benchmark, though scores vary by test setup.
The Decoder · +2 outlets
Meta releases Muse Spark Open, an 816B open-weight MoE model with 1M context
The open-weight release intensifies competition among frontier labs publishing downloadable models with long context windows.
MarkTechPost · +14 outlets
Germany to unveil Nvidia AI agreement next week, government sources say
The planned deal would deepen Germany's ties to Nvidia as European governments push for domestic AI compute capacity.
Reuters · +10 outlets
AI-generated campaign ads flood US midterms as House bill seeks disclosure
Roughly $137 million in AI-generated political ads is flowing into races as lawmakers weigh mandatory AI notices on campaign spots.
The New York Times · +5 outlets
Anthropic gives nonprofit METR inside access to audit its models and safety practices
The arrangement grants an outside evaluator employee-level access, a notable step for third-party scrutiny of frontier AI systems.
The Decoder · +14 outlets
Anthropic co-founder renews calls for robust regulation of frontier AI
Anthropic's leadership is pressing governments for binding transparency and safety standards as the AI arms race accelerates.
Malay Mail · +14 outlets
US AI task force sets path for study and reviews, leader tells CNBC
The federal task force plans to commission a report on AI risks, shaping near-term US policy direction.
Reuters · +4 outlets
G7 to launch exercises to guard against AI cyber attacks
The drills reflect growing concern among major economies about AI-enabled security threats.
Reuters · +11 outlets
IMF chief warns AI risks may exceed opportunities without safeguards
Speaking at the IMF-World Bank meetings, the fund's head called for an ethical framework to accompany AI adoption.
Reuters · +14 outlets
China's DeepSeek hit by five-hour AI service outage
The disruption highlights operational fragility among high-profile Chinese AI providers serving large user bases.
Nikkei Asia · +14 outlets
Google DeepMind's benchmark claims and AI interpretability essay divide researchers
A new essay argues neural networks may be far more legible than assumed, a claim with major implications for AI safety research.
The Decoder
EU pledges safe, sovereign AI but lags far behind US and China
European leaders and Big Tech executives met at an AI summit as the bloc struggles to keep pace in frontier model development.
DW.com · +4 outlets
Senate poised to add Anthropic's Claude to staffers' AI tools
The move would bring a major commercial model into official congressional workflows.
Politico · +2 outlets
Nova Scotia to host massive data centre to power artificial intelligence
The project adds to a wave of large AI compute buildouts seeking cheaper power and land.
CBC · +13 outlets
Taiwan plans $4 billion Texas park to support Nvidia's AI supply chain
The reported plan reflects efforts to onshore parts of the AI hardware supply chain in the United States.
Yahoo Finance
AI's demand for power and water draws scrutiny from utilities and investigators
Growing data centre electricity and cooling needs are reshaping utility planning and local resource debates.
WCTI