US AI task force weighs cyber incident reporting as sharing-law deadline looms
A White House advisory meeting put AI-enabled attacks and information-sharing protections on the agenda, but no new reporting mandate has been announced.
50 stories
A White House advisory meeting put AI-enabled attacks and information-sharing protections on the agenda, but no new reporting mandate has been announced.
Federal Reserve Governor Michael Barr says AI might both help workers and replace tasks; policymakers need to consider the uneven transition.
Google's latest Gemini 4 model aims at practical software and assistant workflows, but a launch announcement is not proof of production performance.
Anthropic's proposed public offering requires careful separation of new shares and shareholder sales. Pricing and final proceeds remain conditional until the official filing and deal terms are verified.
Australia's government has outlined plans for stronger frontier-AI oversight. The proposals are not law, and key details of a reported agent incident remain unresolved.
Open-weight AI can change the economics of model deployment, but API prices and surveys about intentions do not establish real enterprise savings.
Microsoft researchers show how modifying a small number of model components can create a model-integrity threat. Real-world use of this particular method has not been established.
Stanford's AI100 project asks how education, media and other institutions can adapt to rapid AI development. Its report is expert analysis, not a model safety certification.
Google is broadening Gemini 4 Carbon access across its AI offerings. Real production outcomes for long-running agents will depend on reliability, permissions and cost.
UK government-backed investment in Fractile underscores a push for AI inference hardware, but commercial performance, rollout and economics remain to be demonstrated.
Nvidia confirms Jensen Huang has resigned and named CFO Colette Kress interim CEO, with its board searching for a permanent successor.