The US Government Just Ordered an AI Company to Shut Off Its Best Models — And a Tech CEO May Have Triggered It
The AI industry had a genuinely strange week, and it all came to a head on Sunday. A government directive, a CEO conversation that reportedly set things in motion, and a batch of new iPhone features that feel both promising and half-finished — today’s stories share a common thread: AI is now firmly in the territory where governments, executives, and everyday users are all pulling in different directions at once.
Anthropic Was Ordered to Cut Off Global Access to Its Two Most Advanced AI Models
On June 13, the US government issued an export control directive — an order restricting which technologies can leave the country — telling Anthropic to disable its Fable 5 and Mythos 5 models for users everywhere. Anthropic complied immediately, pulling access worldwide. This wasn’t a quiet policy update. Users who logged in found two of the company’s most capable tools simply gone, citing national security concerns.
Think of it like a pharmaceutical company being told by regulators to pull a drug from shelves globally, even in countries that hadn’t asked for it. The concern isn’t necessarily about the product itself — it’s about who might get hold of it and what they might do with it. Export controls have long been used to keep advanced technology out of adversarial hands, and now AI models are being treated the same way.
For everyday Anthropic users, the impact is immediate. If you relied on Fable 5 or Mythos 5 for work — writing, research, coding assistance — you lost access with no prior warning. The Verge reported that the shutdown applied globally, meaning even users in allied countries were cut off.
Why this matters: This is the first time a US government order has forced a major AI lab to disable specific models for all users at once. It signals that Washington now sees advanced AI as a controlled technology, not just a software product.
“US government issued export control directive citing national security concerns.”
An Amazon CEO Conversation With Officials Reportedly Set This Whole Thing in Motion
According to reporting by The Wall Street Journal, Amazon CEO Andy Jassy raised cybersecurity concerns about Anthropic’s AI models with US officials before the government issued its crackdown order. Amazon is a major investor in Anthropic, which puts Jassy in an unusual position: a stakeholder in the company whose models the government then moved to restrict. TechCrunch reported that those conversations appear to have directly influenced the decision.
This matters because it shows how policy decisions about AI aren’t just happening in government offices. A single executive voicing concerns to the right officials can apparently trigger a global access shutdown affecting millions of users. That’s a lot of power concentrated in private conversations most people will never see.
For anyone who uses AI tools at work, the lesson is unsettling. The models you depend on can disappear not because of anything you did, or even because of anything the AI company did, but because of conversations happening far above your pay grade. Business owners and teams building products on top of Anthropic’s API — Application Programming Interface, the connection that lets software talk to AI models — woke up to find their tools simply switched off.
Why this matters: The line between corporate lobbying and government AI regulation is blurrier than most people realized. One CEO’s concerns became global policy within days.
“Amazon CEO Andy Jassy raised cybersecurity concerns regarding Anthropic’s AI models.”
Apple’s New iPhone Photo Tools Are Genuinely Useful, but Don’t Expect Magic
Apple rolled out a set of AI photo editing features with iOS 27, including three main tools: Spatial Reframing (which adjusts the composition of a photo after you’ve taken it), Extend (which fills in the edges of an image to make it wider or taller), and Clean Up (which removes unwanted objects from a photo). As The Verge’s hands-on review noted, the tools work — but they trail behind what Google Pixel phones have offered for some time.
Google has had similar features baked into Pixel devices for years. Apple arriving later means the technology is more mature, but it also means iPhone users are getting a version of something Android users may already take for granted. Clean Up is the most reliable of the three. Extend occasionally produces edges that look slightly off, especially around complex backgrounds like trees or hair.
For the average iPhone user, these are genuinely useful additions, particularly for rescuing photos where someone’s head got cut off or a stranger wandered into the frame. You don’t need any technical knowledge to use them — they show up right inside the Photos app. Just don’t expect them to perform miracles on every image.
Why this matters: Apple’s tools won’t dazzle power users, but they’ll quietly help millions of regular people improve photos they’d otherwise have deleted.
Also Happening in AI
A coalition of 42 state attorneys general, led by New York, has launched an investigation into OpenAI over its advertising practices, according to TechCrunch. India, meanwhile, is watching the Anthropic shutdown closely and using the moment to debate what its own AI strategy should look like, with TechCrunch noting the episode has intensified local conversations about AI sovereignty. Separately, KPMG quietly pulled a report on agentic AI — AI systems designed to take actions on their own rather than just answer questions — after organizations disputed claims in the document, with apparent AI hallucinations (confident but false outputs) blamed for the errors. And in Hollywood, The Verge explored why current AI video tools still can’t produce anything that looks like a real production, despite years of promises.
What to Watch
The Anthropic shutdown is the opening chapter of something larger. Governments now have a working template for ordering AI companies to disable specific models, and other agencies in other countries will be watching to see how the US uses that power. Watch for whether this triggers similar moves in the EU or China — and whether AI companies start lobbying for clearer rules about when and how governments can pull the plug on their products.