Exclusive: Gottheimer presses Anthropic on source code leaks
Add Axios as your preferred source to
see more of our stories on Google.
Photo illustration: Aïda Amer/Axios. Photo: Bill Clark/CQ-Roll Call Inc. via Getty Images
Anthropic needs to explain its recent source code leaks and changes to its internal safety policies, Rep. Josh Gottheimer (D-N.J.) wrote to CEO Dario Amodei on Thursday, warning of potential national security risks in a letter shared exclusively with Axios.
Why it matters: Gottheimer's letter highlights the growing pressure on AI companies from Washington as their tools become embedded in defense and intelligence operations, and comes amid Anthropic's ongoing battle with the Pentagon.
Driving the news: Gottheimer, a leading House Democrat on AI and cybersecurity issues, sent the letter after source material powering Claude Code leaked this week for the second time in over a year.
- "Claude is a critical part of our national security operations. If it is replicated, we sacrifice the competitive edge we have worked so diligently to maintain in all facets of our national security," he wrote.
Zoom in: Gottheimer opened the letter by saying he supports Anthropic in its dealings with the Pentagon.
- "I was deeply concerned with the Trump Administration's decision to block Claude from all government contracts, and how that decision could significantly set us back in our AI race against China."
Gottheimer pressed Amodei on why the company has rolled back certain internal safety protocols.
- He pointed to a Chinese Communist Party-backed group hacking Claude last year, writing that he's concerned upcoming Claude model Mythos will be used to conduct more cyberattacks.
- "Anthropic needs to do everything it can to prevent CCP-backed companies like DeepSeek from conducting distillation campaigns."
Between the lines: Gottheimer is threading a needle — clearly disagreeing with the Trump administration's blacklisting of Anthropic while demanding the company improve its security so advanced AI can still be used in national security operations.
The other side: "A Claude Code release included some internal source code. No sensitive customer data or credentials were involved or exposed," an Anthropic spokesperson told Axios following the leak Tuesday.
- "This was a release packaging issue caused by human error, not a security breach. We're rolling out measures to prevent this from happening again."
What we're watching: Gottheimer asked Anthropic to answer a number of questions about Claude's capabilities, risks of malicious use and susceptibility to further attacks from outside parties.
