#ai-safety
Every summary, chronological. Filter by category, tag, or source from the rail.
Import AI 463: Robotics, Infrastructure, and the Future of Human Agency
This issue covers NVIDIA's new autonomous robotics framework, Tencent's 10,000-GPU diagnostic tools, the historical difficulty of predicting AI's societal impact, and the potential for AI to render human control vestigial.
Moving Beyond Bans: Implementing Safety by Design in the UK
The UK's proposed social media ban for under-16s is a symbolic gesture that fails to address systemic product risks. The incoming government should instead leverage the existing Online Safety Act to mandate 'safety by design'—treating digital platforms like other consumer goods that must be proven safe before release.
Moving Beyond Operation-Level Oversight for AI Agents
Current AI governance fails because it focuses on individual operation approvals rather than the cumulative outcomes of agentic sequences, creating a dangerous gap between human intent and machine execution.
Federal Agencies Mandated to Finalize Quantum-Ready Migration Plans
The OMB has directed federal agencies to finalize post-quantum cryptography (PQC) migration plans within 120 days, setting a phased timeline to secure high-impact systems against future quantum decryption threats by 2035.
FAA Modernizes Air Traffic Control with AI and Data Integration
The FAA has awarded a $876 million, 12-year contract to Air Space Intelligence to build a new data backbone and AI-driven weather visualization tool, aiming to modernize national airspace management amidst rising drone traffic.
Claude Mythos Hits 77.8% SWE-Bench But Stays Gated
Anthropic's Claude Mythos scores 77.8% on SWE-Bench Pro (vs Opus 4.6's 53.4%), finds software vulns like a 27-year-old OpenBSD flaw faster than humans, prompting limited Project Glasswing access to aid patching over public release.
KodeKloudShowing 7 of 7