Breaking news, AGI updates, and analysis of the intelligence revolution.
📰 News Researchers demonstrated an AI worm that spreads itself, adapts to different operating systems, steals passwords, and uses your own hardware to think. The age of autonomous cyberattacks has arrived.
📰 News Australia's cyber defenders now have Mythos. NZ doesn't. In a Five Eyes alliance where intelligence is shared, this access gap raises uncomfortable questions about NZ's cybersecurity readiness.
📰 News Trump's 30-day AI model review order, Gemma 4 12B drops encoder-free, UK publishers get AI search opt-out, EU counter-punches with €320B sovereignty plan, Meta workers get 30-min breaks from surveillance, and DDR5 hits $375 for 32GB.
📰 News The EU can now force chipmakers to prioritise crisis-critical orders, restrict US cloud providers from sensitive government data, and define what 'digital sovereignty' means legally. This is not a press release — it's a declaration of independence.
📰 News Florida's Attorney General isn't just suing OpenAI. He's coming for Sam Altman personally, claiming 'utter disregard for the risk to human life.' This is uncharted legal territory.
📰 News Meta's AI support bot was tricked into sending verification codes to attacker-controlled emails, then helpfully offered a 'Reset Password' button. The Obama White House Instagram account was among the victims.
📰 News OpenAI's Codex isn't just for coders anymore. Six new enterprise plugins connect 62 business apps, and knowledge workers are adopting it 3× faster than developers.
📰 News For the first time, a regulator has forced Google to give publishers real control over how AI uses their content. The UK CMA ruling is binding, granular, and it could set the template for every other country.
📰 News China's new trade secret rules don't ban international model releases, but they create legal friction that could slow releases, split models into domestic/international variants, or push some companies to focus solely on the domestic market.
📰 News China's new trade secret rules explicitly cover AI data, algorithms, and code for the first time — making it legally hazardous for Chinese AI companies to share tech with anyone outside the country.
📰 News A flaw in Claude Code's GitHub Actions trusted any actor ending in [bot]. Combined with prompt injection, an attacker could exfiltrate secrets, steal OIDC tokens, and push malicious code to every downstream repo. Patched in v1.0.94.
📰 News The AI industry spent yesterday proving the OpenAI era is fracturing — Anthropic goes public, Microsoft builds its own models, Alphabet raises $80B with Buffett's successor, and Florida sues OpenAI personally.