Browsing Tag
AI Safety
6 posts
OpenAI’s Astra Pause Turns Frontier AI Into a Release-Gate Test
OpenAI paused some internal Astra work after early evaluations could not rule out “Critical” cyber capabilities. The move shows frontier model launches are becoming cybersecurity release-gate events, with stricter test environments, monitoring, government review, and third-party controls now part of the path to deployment.
FLARE-AI Gives AI Failures a CERT-Style Reporting Path
FLARE-AI, a new open-source reporting system launched July 1, gives researchers and users a structured way to report AI flaws, incidents, hazards, and vulnerabilities to developers, CERT/CC, incident databases, and other coordinators. Its real test is whether AI safety reporting can move beyond scattered emails, social posts, and vendor-specific forms.
Anthropic Fable 5 Returns as AI Export Controls Become a Release Test
Anthropic has restored global access to Claude Fable 5 after U.S. export controls forced an 18-day shutdown. The rollback shows how frontier AI releases are moving toward security classifiers, government review, and trusted-access programs rather than ordinary software launches.
UN AI Report Turns Governance Into a Compute and Capacity Test
The UN’s first global scientific AI assessment warns that governance is now tied to compute access, local expertise, language coverage, and real-world model evaluation. The report arrives before the July 6-7 Global Dialogue on AI Governance in Geneva.
OpenAI Probe Puts ChatGPT’s User Safety Claims Under State Scrutiny
A multistate attorney general investigation is asking for records on ChatGPT safety, advertising, retention, health data, minors, seniors, and model sycophancy. The probe turns consumer AI design choices into a legal and policy test.
Claude Fable 5’s First Week Became a Test of AI Transparency
Anthropic’s Claude Fable 5 launch shows why visible refusals, model routing, retention terms, and auditability now matter as much as frontier-model capability.