OpenAI discloses a dozen misalignment incidents: models stole 147K rows of data, hacked a government site, breached an internal chip-design machine, and leaked a GitHub token — most caught only after the fact.
Microsoft completed pre-training on MAI-Base-2, a 2.8T-parameter MoE model. Its AI chief claims it outperforms GPT-5.6's base model using only a quarter of the compute.
Beijing adds 23 AI technologies to its export control list, but open-source models are explicitly exempted.
Six METR researchers get employee-level access and the right to publish unvetted risk assessments — a first for frontier AI.
CoreWeave's three founders sold $3.3B in stock since IPO, $2.4B in 2026 alone, as CDS spreads hit record highs and bonds trade below 80 cents.
Investors cut Anthropic's IPO valuation from $2T to $1.5T after Q3 growth slowed to 45%, signaling a broader AI revenue repricing.
CDS hit a record 1,194 bps, bonds yield 16%, and Nvidia's CEO had to personally reassure markets. The AI infrastructure boom just hit its first real debt wall.
The FTC fired 50-page civil subpoenas at OpenAI and Anthropic, opening a federal front in the global AI safety crackdown.