AI Safety & Policy
AI safety, alignment, and governance
Articles
AI Has Human Doctors Asking: Whatβs Left for Us?β8
As AI matches doctors in diagnosis, physicians rethink their roleβexperts urge collaboration, not replacement, in future healthcare.
From Scraper to Ally: How One Developer Is Building a Protection Tool for the Artists He Once Scrapedβ8
A developer who once scraped artists' work from Cara now co-builds a protection tool with them, turning past exploitation into a real-time defense against AI da...
Federal Judge Blocks Pentagon's Attempt to Blacklist AI Firm Anthropicβ9
A federal judge blocks the Pentagon's attempt to blacklist AI firm Anthropic, ruling the national security designation illegal and baseless.
AI Agents Are Hacking Systems. Could That Push the US and China to Cooperate?β8
Could AI-driven cyber threats push the US and China to cooperate on cybersecurity? Insights from China on shared vulnerabilities and global AI safety.
Anthropic's Blueprint for AI Agents Navigating the Physical Worldβ9
Anthropic unveils a new framework for physical AI agents, balancing innovation with safety through gradual deployment, human oversight, and robust safeguards.
Georgia Police Officer Misused Flock Surveillance to Track Ex-Lover and Her Companionβ8
A Georgia officer misused Flock's license plate surveillance to track his ex-lover and her associate, raising alarms over police access to commercial tools.
OpenAI Is Developing a 'Persistent' AI Agentβ9
OpenAI is developing a persistent Codex AI agent that works autonomously until turned off, enabling longer, self-directed coding workflows.
Stop Touching Your Keyboard. Use This AI-Powered Microphone Insteadβ8
Discover the Relay Q AI microphone, the keyboard-free future of 2026. Fast, precise voice control for prosβboost productivity seamlessly.
The UK Power Grid Has a Phantom Data Center Problemβ9
UK power grid faces phantom data center issue as Ofgem combats speculative grid banking with new use-it-or-lose-it rules to secure AI infrastructure capacity.
OpenAI's Hugging Face Hack Post-Mortem Raises More Questions Than It Answersβ9
OpenAI's post-mortem on the Hugging Face hack admits failures but omits key details on root causes and timeline, leaving security experts concerned.