AI News Summary — Week 25.1, 2026

This week has been a dramatic one for the AI industry. Here's what's been happening.

US Government Forces Anthropic to Pull Frontier Models

In an unprecedented move, the Trump administration invoked an obscure export control directive to force Anthropic to disable access to its most powerful models — Claude Fable 5 and Mythos 5 — to non-Americans, including the company's own employees.

The Commerce Department sent Anthropic a letter on Friday citing an unspecified national security concern. The company, citing a potential guardrail bypass, complied and shut down both models to all customers worldwide to ensure compliance.

New details emerging over the weekend suggest the real reason may be political rather than technical. Cybersecurity researcher Katie Moussouris published a blog post analyzing the alleged guardrail bypass and concluded that "the bypass itself should never have triggered an export control" — the difference between asking an AI to "review code for security issues" versus "fix this code."

Dozens of top security researchers have since called on the Trump administration to revoke the export control order, calling it "dangerous" and harmful to US cyber defense capabilities.

"The message is that AI companies in the United States can't be trusted to operate without interference from the U.S. government." — Justin Hendrix, Tech Policy Press

Jeff Bezos's Prometheus Raises $12 Billion

Jeff Bezos's physical AI startup Prometheus announced it raised $12 billion** at a staggering **$41 billion valuation. The round included investments from Bezos himself, JPMorgan Chase, Goldman Sachs, and BlackRock.

Prometheus is building what it calls an "artificial general engineer" — software capable of automating the design and manufacturing of complex physical systems, from jet engines to drug compounds. The company currently has 150 employees across San Francisco, London, and Zurich.

Bezos frames the AI revolution differently than many tech leaders: rather than widespread job losses, he predicts "labor scarcity" — a world where demand for human workers outstrips supply thanks to massive productivity gains.

It's the second fundraise round for Prometheus, which launched late last year with an initial raise of $6.2 billion.

OpenAI Files Confidentially for IPO

OpenAI has filed confidentially for an initial public offering, following rival Anthropic's filing just a week earlier. The company submitted a draft registration statement to the SEC for a proposed IPO, though it hasn't decided on timing yet.

"We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company," OpenAI wrote in a blog post. "But it's a complicated set of tradeoffs and this gives us the option to go public sooner if that ends up being best."

With SpaceX also expected to go public at a $1.75 trillion valuation, 2026 is shaping up to be a blockbuster year for tech IPOs.

Meta Brings AI Mode to Facebook

Meta is rolling out new AI features on Facebook, including "AI Mode" — a new search experience that uses Meta AI to surface answers pulled from public posts across the platform, including Groups and Reels. Instead of scrolling through search results, users can ask questions in plain language and get synthesized answers.

The company also introduced AI-powered photo editing tools that let users virtually wear sports team jerseys or try different hairstyles and accessories.

The move follows Meta's quiet launch of Forum, a Reddit-style app with an AI "Ask" tab, last month.

Sources: TechCrunch, The Verge, Axios