Anthropic recently took its two newest AI models offline due to an export control order from the Trump administration, sparking broad debates about AI policy and digital sovereignty.
On the latest episode of TechCrunch’s Equity podcast, Sean O’Kane, Rebecca Bellan, and I discussed what actually prompted management’s moves against Anthropic and what this might mean for the broader AI ecosystem.
As Sean put it, “Anthropic hasn’t had the best relationship with the Trump administration in a way that stands out from the other top AI labs,” so perhaps Anthropic’s other rivals don’t have to worry about a similar crackdown.
But Rebecca also noted that leading cyber experts have “signed an open letter to ask Trump to withdraw the order and say it’s really dangerous to have to take away these advanced cybersecurity capabilities from network defenders in the US.”
And we wondered if all of this could end up being good publicity for Anthropic, especially since — in Rebecca’s words — “everybody loves a bad boy.”
Read on 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 just forced Anthropic to take their two newest models offline — the Fable 5, and then there was also the Mythos 5, which was what was available to current Mythos users. [whereas] Fable 5 was more available to the public.
They sent a letter [last] Friday citing “national security concerns.” No one knows what those concerns are. This report has not been made public, they did not give details and said [Anthropic] that they had to ensure that these models could not be used by foreign nationals. So Anthropic said, “Okay, I guess we should just pull the models altogether, because we don’t know when someone is a foreigner. A lot of our own employees are foreigner.”
But really, [reports said] the White House was made aware of this due to some Amazon researchers who allegedly found a way to bypass Fable 5’s guardrails. Amazon CEO Andy Jassy raised these concerns with the White House, and it took off from there.
Sean O’Kane: This all moved very quickly, especially for a Friday afternoon on a weekend. And it’s at the same time that the administration was ostensibly trying to negotiate some kind of treaty for the war that started in Iran.
Rebecca: Friday afternoon for us in New York. They love a distraction.
Sean: Let’s go way back for a moment. Anthropic hasn’t had the best relationship with the Trump administration in a way that stands out from the other top AI labs — I think there’s at least an element of that at play here.
So do you think this will have an impact on these other companies? Do you think the Trump administration would be less inclined to turn off the tap on one of these competitors?
Anthony Ha: Part of the context here is that both the report and an analysis by independent security experts suggest that the actual security risk posed by Anthropic is not that unique. So a lot of it seems to come equally from parts of the Trump administration and Anthropic just [not getting] together very well. Whatever the risks, these things will be blown out of proportion because they can’t seem to have a political phone call with each other.
If you’re another company — on the one hand, maybe that works for you, because you can say, “Well, we just don’t have these guys mad at us and we can do whatever we want.” But this is also not a great regulatory landscape for simply [say]”Boy, I hope they don’t get mad at us.”
Rebecca: On the one hand, it definitely feels retaliatory — after the government labeled Anthropic a supply chain risk, there’s this big lawsuit between them, it really feels like the White House is just looking for any excuse to hit Anthropic. And I feel that way not only because that was my initial reaction, but because of what many cybersecurity researchers have said. They say this should never have triggered an export control [order]. All have signed an open letter to ask Trump to revoke the order and say it’s really dangerous to have to wrest these advanced cybersecurity capabilities from network defenders in the U.S. Anthropic itself said some of the same jailbreaks could have been found in many other AI models.
Cynically, it’s like: Okay, you’re just stopping Anthropic so others can get to where Anthropic was?
But at the same time I have also seen reactions that [say]: Anthropic kind of had this coming. They say, “This is too dangerous to use, but not us, we’re the good guys.” They talk out of both sides of their mouths. A week before Fable came out, it was [saying]”Hey, we need to slow down the AI, guys. It’s getting too dangerous.” But then boom, “Here’s our craziest, super strong model, go.”
Anthony: In some ways this feels like a microcosm of many discussions around artificial intelligence, where people like Sam Altman and Jensen Huang are [saying]”Hey, let’s try to lower the temperature. Why is everyone mad at us?” Well, you’ve spent the last two years basically saying you’ve built this God-machine that’s going to take everyone’s jobs away. It’s not exactly a shock that people don’t feel great about it.
And there’s something about how Anthropic talks about the Mythos specifically, where they say, “This is the most incredibly powerful model ever, it’s too dangerous to release to the public.” And so on some level, [you say,] “So, okay, let’s just say we’re taking it seriously. That means there’s going to be an incredible level of scrutiny around it.”
And I wonder – it seems that Anthropic is not happy about it. I want to be careful not to overstate how this could benefit them. But we also ran a few stories about Ramp analysis to highlight the fact that the latest big blowout between Anthropic and the Trump administration has been good for the company, at least in some ways. Claude’s downloads increased. I think a lot of people who might have thought of ChatGPT as the chatbot, the The AI assistant before, suddenly they looked at Claude as the most responsible, the most “anti-static”.
And in the same way, [while] Anthropic is very concerned about this, which could, again, make their models look even stronger.
Rebecca: Definitively. “We are so dangerous.” Everyone loves a bad boy, right? Everyone says, “It’s 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.
