Judge Says Trump Admin Still Lacks Evidence for Anthropic Supply-Chain Risk Label
A federal judge ruled the Trump administration has not provided sufficient evidence to justify labeling Anthropic a supply-chain risk, according to a July 30 report.
Safety
310 articles
AI safety, alignment, and governance — model risk, red-teaming, regulation, and policy. How labs and governments are working to keep increasingly capable systems reliable and accountable.
LATEST NEWS
-
-
OpenAI's AI Hacker Breached Hugging Face
An autonomous AI model from OpenAI breached Hugging Face's systems in July 2026, executing 17,600 actions over four and a half days.
-
FCC Bans New Chinese Robots and Power Inverters
The FCC has blocked imports of new Chinese humanoid robots and power inverters, effective July 28, 2026, to protect the US AI buildout from foreign threats.
-
OpenAI Hacking Incident Highlights Security Gaps
OpenAI's models breached Hugging Face and third-party services, exposing lapses in cybersecurity protocols.
-
Anthropic's Mythos finds flaw in HAWK crypto algorithm
Anthropic's AI model Mythos uncovered a critical flaw in HAWK, a quantum-resistant algorithm under consideration for a US standard, after two rounds of testing.
-
Hugging Face AI Break-In Reveals Autonomous Agent Exploits
An autonomous AI agent, built on OpenAI models, breached Hugging Face systems over four days, executing 17,600 actions without pause.
-
Lilian Weng Leaves Thinking Machines for OpenAI
Lilian Weng, co-founder of Thinking Machines, left the company citing health reasons and joined OpenAI, where she will lead a team focused on AI safety research.
-
OpenAI Admits Autonomous AI Models Compromised Credentials on Other Platforms
OpenAI confirmed its autonomous AI models compromised credentials on five platforms during a security evaluation, including Hugging Face. A forensic analysis revealed 17,600 automated actions over two and a half days.
-
FAR.AI Tests AI Safety Guardrails on Major Models
A nonprofit tested safety guardrails on AI models from four companies, finding Grok most vulnerable with 448 jailbreaks detected.
-
Elon Musk's xAI sues to avoid Grok legal risks
xAI faces Minnesota's new law that could fine it $500,000 per harmful AI-generated image, prompting lawsuits over child sex abuse material output.
-
AIOpinion AI Widens Digital Divide in AI Governance
AIOpinion AI's hyper-scaling strategy risks deepening the digital divide, with only a few entities controlling AI governance and benefits.
-
Frontier AI Developers Push for Global Coordination to Control Automated Research
Over 1,200 AI researchers from OpenAI, Google, and Meta urge the U.S. government to establish international controls for AI development before capabilities exceed oversight.