Anthropic’s METR Access Opens a New Test of Independent Scrutiny
Embedded evaluators may see more, but access limits and publication terms will determine what the public can learn from their findings.
Share
Policy, Safety & Society
327 posts
AI regulation, government, safety, security, geopolitics, copyright, misinformation, privacy, human rights and the changing world of work.
Flock’s Published Audio Policy Reaches Beyond Gunfire
Flock's technical explanation distinguishes event-triggered clips, incidental speech and multiple sound categories, creating concrete questions about public-space audio sensing.
Share
Harvard’s AI Gender Study Separates Confidence From Knowledge
A new Shorenstein Center study distinguishes uncertainty from AI knowledge and connects participation with workplace support and public visibility.
Share
New Terrorism Benchmark Puts Modified AI Safeguards Under Scrutiny
New reporting on modified AI models highlights the limits of safety scores, while earlier public research clarifies what counterterrorism evaluations can establish.
Share
Amex’s Purchase-Intent Controls Put Boundaries Around Agent Payments
Public ACE documentation separates customer consent, scoped credentials, and planned protections, with important limits on what can be verified today.
Share
AI Campaign Spending Puts the Identity of Political Vendors in Focus
New reporting on AI-related political spending renews a contested disclosure question, while an earlier complaint and FEC guidance clarify what is alleged.
Share
AISI Wants Frontier AI Oversight to Begin Before Public Release
AISI's new technical agenda extends oversight across development and deployment while distinguishing voluntary best practice from mandatory requirements.
Share
Britain’s AI Safety Push Tests the Gap Between Regulation and Research
Current British AI measures and Microsoft's published governance proposals show why technical access, voluntary standards and enforceable authority are different commitments.
Share
Ambi’s Privacy Promises Meet More Detailed Audio Retention Rules
Ambi's privacy and service documents distinguish listening audio, saved recordings, derived memory and authorization for external actions.
Share
Kyvvu’s Agent Security Pitch Depends on the Whole Task, Not One Permission
Kyvvu's action-control approach follows task history, while its documentation shows why policy design and complete integration remain central to the security claim.
Share