Amazon CEO triggered the Anthropic shutdown; 42 states probe OpenAI

A Wall Street Journal report ties the government's abrupt shutdown of Anthropic's Fable 5 and Mythos models to Amazon CEO Andy Jassy, who reportedly warned Treasury Secretary Scott Bessent that Amazon researchers had used Fable 5 to gather information useful for cyberattacks. The twist: Amazon is one of Anthropic's largest investors, with a 5 billion dollar cloud commitment, yet also a direct rival. David Sacks, the administration's former AI czar, said a trusted partner flagged the flaw and that Anthropic declined to fix it. Anthropic calls the order a possible misunderstanding and says it expects access to be restored.

A coalition of 42 state attorneys general has opened a sweeping investigation into OpenAI, with New York's office issuing a subpoena on June 12. The demand seeks documents on OpenAI's advertising, its user-engagement and retention tactics, its handling of consumer and health data, model sycophancy, and its treatment of minors and seniors. The probe lands days after OpenAI confidentially filed to go public, adding regulatory risk to its IPO and threatening to force changes in how ChatGPT is built and marketed.

Meta has begun unwinding its 2 billion dollar acquisition of Manus, the Chinese-founded AI agent startup, after China's National Development and Reform Commission ordered the deal reversed on national-security grounds in late April. Meta has cut Manus off from its internal systems and halted data sharing as the two separate. Manus's founders are reportedly seeking about 1 billion dollars to buy the company back, a path that could lead to a Chinese joint venture and a Hong Kong listing. The episode shows Beijing will block offshore deals for strategically sensitive AI, wherever the company is incorporated.

The same directive that pulled Anthropic's Fable 5 and Mythos has reopened a hard question in India, Anthropic's second-largest market: how much of its AI future should it rent from foreign labs? The suspension hit just after Tata Consultancy Services began training 50,000 staff on Anthropic's models. Sarvam CEO Pratyush Kumar argued that India should "not confuse access with ownership," pointing to his firm's home-grown 105 billion-parameter model and its own GPU clusters as the sovereign alternative, while Aarin Capital's Mohandas Pai called for a national AI fund of roughly 6 billion dollars. Expect louder calls for open-weight adoption.

China's Zhipu AI released GLM 5.2, a coding-first model with a 1 million token context window, pushing it live across all four tiers of its GLM Coding Plan. The company is positioning it as a permissively licensed alternative to Claude Code and GPT-5.5 for the Asia-Pacific market, with a standalone API and MIT-licensed open weights promised for next week. One caveat: Zhipu published no benchmarks at launch — no SWE-bench, no LiveCodeBench — so its "powerful coding" and "long-horizon" claims stay unverified for now, echoing the skepticism that met Moonshot's Kimi release a day earlier.

KPMG has pulled a report titled "Redefining excellence in the age of agentic AI," published in October 2025, after the organizations it praised said the claims were made up. UBS, the UK's National Health Service, Swiss Federal Railways, and Transport for London told the Financial Times that the report's descriptions of their AI use were untrue or misleading, and detection firm GPTZero traced the errors to AI hallucinations; a forensic review found most of its source citations were mangled, fabricated, or too vague to verify. The irony writes itself — a Big Four firm that sells AI-governance services shipped an AI-written report that hallucinated its own evidence, weeks after a similar EY retraction.

Amid the running debate over AI's energy and hardware footprint, researchers at UC San Diego, backed by Google, turned discarded smartphones into low-carbon servers. By stripping phones down to their motherboards — which hold roughly half of a device's embedded carbon — and running Linux and Kubernetes across clusters of 25 to 50 units, a 20-phone rack matched a standard cloud server's latency for a class of more than 75 students. The team plans a 2,000-phone cluster later in 2026 that should deliver about 50 servers' worth of compute at a fraction of the usual cost, extending device lifespans instead of mining new materials.