AI leaves the lab. We follow what happens next.

Live Wire·Archive September 13, 2026 Daily Briefing
Listen to today's briefing 18:46
Signal: The warnings stopped being talk: lawmakers reached for criminal penalties, Microsoft wrote its own rules, and the money kept building anyway.
California Drafts Emergency AI Bills as Khanna Demands Charges
California lawmakers want emergency legislation and criminal penalties for rogue AI, with Ro Khanna calling Amodei's pacing plan nowhere near enough and demanding a ban on self-improving models - a day after Trump dismissed the warnings as things that won't happen.
Nadella: Microsoft publishes its own MAI model rules Monday, and not from three labs Lina Khan: firms must face liability for dangerous, unvetted products NY Post: Trump calls the warnings 'things that won't happen'

Xi Sells BRICS on Open AI as Beijing Weighs Curbs
Xi offered BRICS a China-led open-source AI community and model cooperation, yet the bloc's declaration ignored it and Beijing spent the summer debating whether to cut foreign access to Qwen, Doubao and GLM, an openness with a ceiling nobody has written down.
CNBC: Xi's statement never mentions the safety debate Fortune: Altman says a US-China AI pact would win Trump and Xi the Nobel
AI Can Unmask Georgia Ballots Before November
A known Dominion flaw lets shuffled ballot records be put back in order, and a Princeton researcher says a public AI assistant did the unshuffling and matched many voters to ballots from May's primary - Georgia says it has mitigated, advocates say not enough.
Obama Tells Democrats to Make AI Their Fight
Obama told a private fundraiser that AI 'moving very fast in private hands' could be dangerous and urged Jeffries to build a public framework if Democrats win the House, making safety a 2028 platform plank; the quotes come from a partial transcript.
ABC: Jeffries says AI tops Democrats' Tuesday caucus agenda
Ghost Note The Cipher Held 370 Years. Fable Took 44 Minutes.