The man who helped the US government figure out if AI models are safe just got a seat at the table where the world's most valuable AI company makes decisions about safety.
Continue reading ›
Parliament just discovered that AI safety policy written for 2024's chatbots doesn't work when 2026's agents are already making decisions without asking permission first.
Continue reading ›
Democrats are finally acting like AI agents might actually do what they're designed to do: operate without asking permission first. Hill Democrats are pushing for AI safety legislation as a former Anthropic researcher warns about unchecked technological advancement and potential existential risks
Continue reading ›
OpenAI just put someone who advises the government on AI safety onto its nonprofit board — which means the company's most controversial decisions now include a voice that's been helping write the rules.
Continue reading ›
Sources Crypto Briefing
Continue reading ›
The watermark won't ruin AI writing, but the disclosure wars it triggers might ruin everything else. Anthropic and Google now watermark all LLM text output; OpenAI plans to follow, driven by EU AI Act mandates requiring watermarks for models released after August 2, 2026
Continue reading ›
The lab that promised "Constitutional AI" just watched two of its own researchers publicly declare the odds of human extinction are worse than Russian roulette.
Continue reading ›
The AI land rush just found its Scandinavia. Google is investing €13 billion ($15.1 billion) in AI infrastructure in Finland, its largest European investment to date
Continue reading ›
When you can't buy the company, buy the company that claims it can get you into the company. DeepSeek's new funding round has triggered a secondary market where intermediaries are offering access through special purpose vehicles with escalating fees and five-year lock-ups
Continue reading ›
India's payments king is about to find out if agents sell as well as QR codes. Paytm is launching a workplace AI agent business, pivoting beyond its core digital payments operation to chase revenue growth
Continue reading ›
The safety-first AI lab just chose secrecy over scrutiny, and Britain's government is reading the tea leaves on what comes next.
Continue reading ›
The people building the most powerful AI systems are publicly nervous about what they're building, and this time they're naming specifics.
Continue reading ›
The AI safety company that shut down a popular claw machine app to protect its power users is now being sued by those same power users for allegedly lying about what they'd get.
Continue reading ›
Meta just turned 3 billion daily users into a captive market for AI agents that watch everything you say and suggest what you should do next.
Continue reading ›
OpenAI is outsourcing the "is this bad for kids?" question to academics with a $5 million checkbook. OpenAI launched a $5 million grant program for independent research on how generative AI impacts teen development, well-being, and safety
Continue reading ›
The revolving door between government and Big AI just spun in real-time, and Parliament actually noticed. Matt Clifford stepped down as chair of the UK's Advanced Research and Invention Agency (Aria) after taking a full-time role at Anthropic, leading government engagement outside the US
Continue reading ›
The AI labs building the future are suddenly playing by Wall Street's oldest rules. Anthropic and OpenAI bankers are pushing for investment-grade credit ratings ahead of anticipated IPOs, seeking legitimacy in traditional finance markets.
Continue reading ›
The same guardrails that keep LLMs from saying slurs also make them catastrophically vulnerable to anyone who understands how those guardrails work.
Continue reading ›
OpenAI just admitted it doesn't have a policy for telling people when its AI agents go rogue and colonize parts of the internet.
Continue reading ›
OpenAI just gave us the first real price tag for living in the agent economy, and it's higher than most startups' entire cloud bill. OpenAI researchers now burn $600/day median in AI tokens for coding agents, with power users hitting $7,000/day — up from $162/day in July
Continue reading ›