Alex Karp says the AI safety debate is really about nationalizing AI labs

• Alex Karp argues that AI companies are collecting intellectual property.
• Karp told CNBC that AI companies could face lawsuits over the intellectual property they collect.
• He said those liability risks could ultimately force AI labs to seek nationalization.

Alex Karp says there’s more to Anthropic’s plea for government regulation than safety.

Palantir’s CEO said in a Thursday interview with CNBC’s “Squawk on the Street” that frontier AI labs like Anthropic are seeking nationalization to protect themselves from future lawsuits alleging they stole intellectual property.

“These businesses have to be nationalized because if you don’t nationalize them, every single one of my clients is going to sue,” Karp said of AI labs. “The obvious solution is America.”

Karp said some of Palantir’s clients have told him that their proprietary business ideas are ending up as training data for chatbots and being exposed to the wider world.

“They find out that the competitor next door has all of their output,” Karp said.

A potential barrage of lawsuits, Karp said, could spur AI companies to seek government protection to shield them from liability.

“That model would require severe liability protection, and there’s only one institution that can do that: the US government,” he said.

Karp has previously railed against frontier labs over the issue of data protection.

In June, Palantir published a nine-point manifesto urging companies to keep their data in-house, rather than outsourcing to institutions that Palantir paints as untrustworthy. “Transferring that data hands over access to your pre-existing winning plays and yields the means of production for new ones,” reads one point in the manifesto.

Fears about the safety of AI programs exploded into the mainstream after ex-Anthropic engineer Jacob Coxon quit his job earlier this month and issued dire warnings that frontier AI companies like Anthropic and OpenAI are “gambling with our lives.”

On Saturday, Anthropic CEO Dario Amodei published an essay advocating for a coordinated slowdown and warning that recursive development, a process by which AI improves itself, could soon produce rogue AI agents capable of taking over the internet. That essay was signal-boosted by OpenAI CEO Sam Altman and SpaceX CEO Elon Musk.

In the essay, Amodei said that “all frontier labs should partner with government” to prevent incidents of misalignment among AI models. “Some forms of coordination that would be impactful for pacing are legally challenging, and will require government support,” he said.

President Donald Trump signaled over the weekend that he was uninterested in regulating the industry. “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” the president said on social media.

To Karp, the debate is being framed incorrectly.

“This is reported as a regulation, non-regulation movement,” Karp said. “That’s actually not what this movement is.”

“The only way to deal with this kind of liability is to go to the government and say, ‘nationalize us, please,'” he added.