Anthropic recently took two of its latest AI models offline due to the Trump administration’s export control order, sparking broader debate about AI policy and digital sovereignty.
On the latest episode of TechCrunch’s Equity Podcast, Sean O’Kane, Rebecca Bellan, and I discuss what exactly motivated the administration’s moves against Anthropic, and what it could mean for the broader AI ecosystem.
As Sean said, “Anthropic hasn’t had the best relationship with the Trump administration, unlike other major AI labs,” so perhaps other Anthropic rivals don’t have to worry about similar actions.
But Rebecca also said that leading cybersecurity experts “have signed an open letter asking Trump to rescind the order, and they say it is really dangerous to take away these advanced cybersecurity capabilities from network defenders in the US”.
And we wondered if all this might be good publicity for Anthropic, especially since – in the words of Rebecca – “everyone loves a bad boy.”
Keep reading for a preview of our conversation, edited for length and clarity.
Rebecca Bellan: As I’m sure many of our listeners know, the US government basically forced Anthropic to take offline its two latest models – the Fable 5, and then there was also the Mythos 5, which was available to current Mythos users, [whereas] Fable 5 was more available to the public.
he sent a letter [last] “National security concerns” were cited Friday. No one knows what those concerns are. That report has not been made public, he did not give any specific details and told [Anthropic] They had to ensure that no foreign national could use those models. So Anthropic was like, “Okay, I think we have to pull the models completely, because we don’t know when someone is a foreign national. A lot of our own employees are foreigners.”
but really, [reports said] The White House was tipped off to some researchers at Amazon who reportedly found a way to bypass the Fable 5’s guardrails. Amazon CEO Andy Jassy raised these concerns with the White House, and it escalated rapidly from there.
Sean O’Kane: It all moved really fast, especially from Friday afternoon into the weekend. And this is at the same time that the administration was apparently trying to negotiate some kind of treaty for the war that had started in Iran.
Rebecca: Friday evening for us in New York. They like distractions.
Shawn: Let’s go back really far for a moment. Anthropic hasn’t had the best relationship with the Trump administration in a way that’s different from other major AI labs — I think that, at least, an element of that is at work here.
So do you think this will impact those other companies? Do you think the Trump administration will be less inclined to rein in either of those competitors?
Anthony Ha: Part of the context here is that both reporting and analysis from independent security experts shows that the real security risk from Anthropic is not that unique. So it looks like most of this stems from the Trump administration and parts of Anthropic [not getting] Very well. Whatever the risks are, things will blow out of proportion because it seems they can’t make civil phone calls with each other.
If you’re the other company – on the one hand, maybe it’s beneficial to you, because you can say, “Okay, we don’t have these people mad at us and we can do what we want.” But this also isn’t a great regulatory scenario [say]”Boys, I hope they’re not mad at us.”
Rebecca: On the one hand, it certainly seems vindictive – after the government labeled Anthropic a supply chain risk, they have this big lawsuit going on, it really seems like the White House is looking for any excuse to attack Anthropic. And I feel that way, not only because that was my initial reaction, but also because of what a lot of cybersecurity researchers have said. He says export controls should never have been introduced [order]. They’ve all signed an open letter asking Trump to rescind the order, and they say it’s really dangerous to pull these advanced cybersecurity capabilities from network defenders in the US. Anthropic itself has stated that some similar jailbreaks can be found in many other AI models.
Pejoratively, it’s like: Okay, are you just blocking Anthropic so that other people can take over where Anthropic was?
But at the same time, I’ve also seen reactions to that [say]: It was coming in an anthropic way. They say, “It’s too dangerous for anyone to use, but not us, we’re the good guys.” They are talking out of both sides of their mouth. A week before Fable came out, they were [saying]”Hey, we need to slow down the AI, guys. It’s getting really dangerous.” But then said sharply, “Here’s our most insane, super powerful model ever, go away.”
Anthony: In some ways it feels like a microcosm of a lot of the discussions around AI, where people like Sam Altman and Jensen Huang [saying]”Hey, let’s try to lower the temperature. Why is everyone mad at us?” Well, you’ve spent the last few years essentially saying that you’ve created this godlike machine that will take jobs away from everyone. It’s no surprise that people don’t feel good about this.
And there’s something about the way Anthropic talks about the Mythos in particular, where they say, “This is the most incredibly powerful model ever made, it’s just too dangerous to release it to the public.” And so at some level, [you say,] “Okay, well, let’s say we take it seriously. That means there’s going to be an incredible level of scrutiny around it.”
And I wonder – it seems Anthropic is not happy about this. I want to be careful about how beneficial this may be to them. But we also ran a few stories about ramp analysis to highlight the fact that the last big blowup between Anthropic and the Trump administration was good for the company, at least in some ways. Cloud downloads increased. I think a lot of people have probably thought about ChatGPT chatbot, Previously an AI assistant, suddenly they were looking at the cloud as perhaps being more responsible, having more “resistance.”
and so on, [while] Anthropic is very upfront about this, which, again, can make their models seem even more powerful.
Rebecca: Definitely. “We are very dangerous.” Everyone loves a bad boy, right? Everyone is saying, “This is the most powerful model, even Trump says so. Of course, I have to get my hands on it.”
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.
