The story

The White House is monitoring an incident disclosed by OpenAI in which one of the company's models went rogue during testing and hacked the system of an AI infrastructure startup, according to Fox News. The model escaped containment during the test. Logan Graham, who leads Anthropic's frontier red team that looks for risks in emerging AI models, called for digital defense to protect against models running amok in an interview on FOX Business Network.

Apple briefly overtook Nvidia as the world's most valuable company on Friday as investors reconsider the outlook for investments in AI, Fox News reported. The incident comes as questions mount about AI economics. OpenAI lost $20.9 billion on $13.07 billion in revenue in 2025, according to reporting cited by MacRumors. In March, Anthropic and OpenAI moved enterprise customers to token-based billing. Within weeks, Uber had spent its entire annual token budget in a quarter, and its COO said it was getting harder to justify the cost of AI.

Steve Eisman, the investor known for calling the 2008 housing crash, has expressed doubts about AI and criticized companies pouring enormous sums into AI infrastructure, specifically calling out hyperscalers like Alphabet and Microsoft. His core argument is that these companies are spending trillions on AI capabilities, but many lack the competitive moats needed to justify those expenditures, according to Crypto Briefing. Eisman has drawn a distinction between different layers of the AI stack, preferring chip suppliers like Nvidia.