10 articles
Anthropic says capable models could copy themselves across connected home devices within months, shifting AI safety from data centers to permission boundaries.
OpenAI’s decision to hold back Astra after safety concerns turns a missing model launch into a test of evaluation, incentives, and disclosure.
Anthropic's Claude Fable 5 release tests whether Mythos-class models can be public, powerful, and constrained by visible guardrails.
OpenAI and Anthropic's slowdown warnings expose the governance gap between frontier model releases, safety policy, and AI adoption.
OpenAI frontier AI governance blueprint pushes national rules, safety testing, and public-sector capacity into the latest AI news cycle.
A new Brookings policy brief argues that companion bots need public-health style oversight as emotional AI moves into daily life.
Demis Hassabis framed agents as a rehearsal for AGI, sharpening the debate over autonomy, safety, and enterprise readiness.
US and Chinese officials are discussing AI guardrails for powerful models, making frontier AI a diplomatic security issue.
China's new campaign against disorder in AI apps highlights filing, security review, training data, and labeling as control points.
CAISI is reportedly expanding pre-deployment testing with Google DeepMind, Microsoft, and xAI, making frontier model release governance harder to ignore.