US pulls Anthropic's top models offline; Bezos's Prometheus raises $12 billion

On June 12, the US government issued an unprecedented export-control directive ordering Anthropic to block all foreign nationals — whether inside or outside the country — from its two most capable models, Fable 5 and Mythos 5, citing national security. Because the company cannot reliably tell foreign users apart from everyone else in real time, it took both models fully offline worldwide while it works to restore access. Anthropic is complying but openly disagreed, warning that the cited jailbreak is a capability already present in rival models and that applying this recall standard "would essentially halt all new model deployments for all frontier model providers."

Prometheus, the physical-AI startup co-led by Jeff Bezos and former Google scientist Vik Bajaj, raised $12 billion in a Series B round that valued it at $41 billion. The company is building what it calls an "artificial general engineer" — software trained on real-world experimental and robotics data to automate the design and manufacturing of complex physical systems, from jet engines to drug compounds. With backers including JPMorgan, BlackRock, and Goldman Sachs, the roughly seven-month-old company has now raised more than $18 billion.

EngineAI, a Shenzhen startup that began shipping its T800 humanoid robots this month, confidentially filed for a Hong Kong IPO — the latest entrant in a fast-widening race to take China's robot makers public. Sector leader Unitree has filed for a $7 billion listing, hand-maker PaXini and robot-vacuum giant Dreame are weighing debuts, and robot-hand unicorn Linkerbot is chasing a $6 billion valuation. The rush reflects how much capital has flooded Chinese embodied AI, with backers spanning Alibaba, Tencent, CATL, and Geely.

OpenAI rolled out a rebuilt ChatGPT memory system to Plus and Pro subscribers in the US, doubling memory capacity and letting the assistant update what it remembers automatically over time. OpenAI says the new design sharply improves factual recall and how closely ChatGPT follows a user's stated preferences as conversations pile up across weeks and months. The upgrade landed the same week that OpenAI retired its older GPT-5.2 models from ChatGPT.

China's Moonshot AI released Kimi K2.7 Code, a one-trillion-parameter open-weights coding model that the lab says cuts reasoning-token use by about 30 percent while topping its previous K2.6 release on internal coding benchmarks. Moonshot published the full model weights to Hugging Face the same day and priced it well below Western flagships. But practitioners quoted by VentureBeat cautioned that the headline benchmark gains have been hard to reproduce in real-world use — a recurring pattern as open-source labs race to claim coding leadership.