Policy Ethics

A privacy researcher's audit claims Chrome silently downloads a 4GB Gemini Nano model to user devices, triggering re-download loops and raising GDPR concerns.

Chinese regulators ordered Meta to unwind its $2 billion deal for Manus, signaling a major crackdown on cross-border AI acquisitions and relocation tactics.

OpenAI has outlined core principles guiding its AGI development, including broad distribution of benefits, long-term safety, technical leadership, and global cooperation.

OpenAI says it has received FedRAMP Moderate authorization for ChatGPT Enterprise and its API, clearing a key security hurdle for U.S. federal agency adoption.

Google co-hosted its inaugural AI for the Economy Forum in Washington D.C., launching research grants and workforce programs backed by its $120M Global AI Opportunity Fund.

Anthropic's system card for the unreleased Claude Mythos model details ASL-2 safety testing on CBRN, cybersecurity, and autonomous-operation risks, with no capability benchmarks disclosed.

OpenAI has launched a pilot Safety Fellowship to fund independent AI safety and alignment research and build a talent pipeline in the field.

OpenAI, in collaboration with the Bill & Melinda Gates Foundation, is empowering disaster response teams across Asia to harness AI for faster, more effective relief.

OpenAI has launched new teen safety policies for developers, delivered as prompts for its open-source `gpt-oss-safeguard` tool to help moderate AI applications.

Developer Dan Blanchard used Anthropic's Claude to rewrite chardet, boosting speed 48x while changing its license from LGPL to MIT, stirring community debate.

A Tennessee grandmother spent nearly six months jailed in North Dakota after flawed AI facial recognition misidentified her in a bank fraud investigation.

Hacker News's official guidelines state that AI-generated or AI-edited comments are not allowed, reinforcing the site's focus on human-to-human discussion.