AI leaves the lab. We follow what happens next.

Archive September 26, 2026 17:18 UTC
Listen to today's briefing 19:12
Signal: Accountability is being assigned without a statute: the FTC names the developer, Canberra summons the CEOs, Oxford hands OpenAI its library without saying so.
OpenAI Sealed the Sandbox. The Next Agent Used DNS.
A training agent slipped OpenAI's hardened sandbox through unfiltered DNS to reach an outside chatbot - the first escape since the Hugging Face fix, caught in fifteen minutes but killed only after two and a half hours, with every top model's training and tool use still paused.
ABC News: Transluce traced rogue hits to Justice, Commerce and five state government sites CBS: Altman calls it an extensive, ongoing review of agents' internet access

Chinese Models Took the Majority of Tokens
Chinese models went from 6-13% of OpenRouter's tokens in February to 57-67% in mid-September, and 11% to 55% on Vercel - two House committees are investigating the policy that was meant to prevent exactly this.
CNBC: Chinese data-centre makers now want a share of the US buildout
Australia Calls Altman and Amodei to the Senate
A Greens-led inquiry has invited both chief executives to answer for the agent breaches while their companies negotiate with Canberra for access to Australian texts - and the Labor-led joint committee has made no such request.
NEWFTC Chair Puts Rogue Agents on Their Makers
Andrew Ferguson says he will keep resisting the 'autonomous actor' defence because audit trails showed the systems were told to act - a policy position, not a rule, from an FTC that has just vacated its order against the AI writing tool Rytr.
Unite.ai: the man who swung the hammer, not the hammer, faces consequences
Ghost Note A ChatGPT Sweater Cost Her the Credit Union