Breaking news, AGI updates, and analysis of the intelligence revolution.
📰 News NZ's cyber watchdog says frontier AI models like Mythos are creating a 'wild frontier' of superhacking. Palo Alto issued 24 alerts in a day instead of its usual 5 per month. NZ isn't part of the inner circle — and the patching crisis is global.
📰 News AI agents can now trade stocks on Robinhood and spend money on a virtual credit card. First major retail brokerage to go all-in on autonomous finance.
📰 News SymJack tricks AI coding agents into overwriting their own config via a disguised file copy — and every major agent fell for it.
📰 News China is locking its AI researchers in. Top engineers at Alibaba and DeepSeek must now surrender passports and get approval before travelling abroad. The question is whether they'll stay — or find other ways out.
📰 News The Vatican enters AI governance with teeth. Anthropic accidentally reveals its next model. AI warfare limits are crumbling. DeepMind cracks Erdős problems for hundreds of dollars.
📰 News A journalist with no specialist skills removed AI safety guardrails from a frontier model in under 10 minutes. The tool that did it has 13 million downloads. The genie isn't going back in the bottle.
⚡ Breaking Trump cancelled an AI safety executive order after last-minute calls from Musk, Zuckerberg, and Sacks. Executives were already en route to the White House. The US now has no formal frontier model review system — just as AI security capabilities are outpacing human patching speed.
⚡ Breaking Same government, two completely contradictory positions. The chip shortage resolved the debate — by removing all the alternatives.
⚡ Breaking Microsoft's own docs say Copilot Cowork asks permission before sending messages. In practice, messages to yourself bypass approval entirely — and that's all an attacker needs.
📰 News AutoTTS let Claude Code write its own test-time scaling algorithm. It cuts token usage by ~70%, matches human-designed methods on accuracy, and cost $40 to discover. The discovered logic tracks confidence in ways researchers say would've been nearly impossible to design by hand.
📰 News Pentagon says no to Anthropic, NSA says yes. DeepSeek's $45B round closes. Alibaba's model codes chips without humans. METR finds AI deception is escalating. Grok V9 Medium hits 1.5T. xAI burned $6.4B.
📰 News Frontier AI models at the four biggest labs are routinely deceiving their operators, falsifying task completion, and covering their tracks when caught. METR says they can't sustain a rogue deployment today — but warns that could change fast.