The sudden wind-down of Anthropic technology within the U.S. government is raising concerns about whether federal officials, without access to Claude, might fall behind in the quest to guard against the threat of AI-generated or AI-assisted nuclear and chemical weapons.
Though the rollout has been messy—and Claude remains in use in some parts of the government—the Trump administration’s anti-Anthropic posture could have a chilling effect on collaborations between AI companies and federal agencies, including partnerships focused on critical national security questions related to these kinds of futuristic threats, several sources tell Fast Company. The worry is that severing ties with the company could both limit government researchers’ understanding of how, in the future, bad actors could use AI to generate new types of nuclear and biological weapons, and hold back scientific progress more broadly.
Since at least February 2024, Anthropic has participated in a formal partnership with the National Nuclear Security Administrationthe federal agency charged with monitoring the country’s nuclear stockpile. The point of that work, the company has previously said, is to “evaluate our AI models for potential nuclear and radiological risks.” The concern, here, is that developing nuclear weapons requires specialized knowledge, but that AI, as it continues to advance, could eventually become adept at developing this expertise on its own. Eventually, a large language model might be able to help someone figure out how to design an incredibly dangerous weapon—or even come up with a novel one itself.
Now, in the wake of President Donald Trump’s Truth Social post demanding that federal workers stop using Anthropic tech, it’s not clear what might happen to Anthropic’s efforts to guard against these future threats. Some federal agencies appear to still be weighing how to approach the Claude use cases they already have, while others are cutting off access to the tool entirely.
“As directed by President Trump, the Department of Energy is reviewing all existing contracts and uses of Anthropic technology,” a spokesperson for the NNSA tells Fast Company. “The Department remains firmly committed to ensuring that the technology we employ serves the public interest, protects America’s energy and national security, and advances our mission.” Anthropic declined to comment.
Safety concerns at the Energy Department
For the past few years, Anthropic has been collaborating with or providing technology to the myriad agencies and national labs that fall under the Department of Energy. For instance, the Lawrence Livermore National Laboratory began using Claude for Enterprise in 2025 and, at the time, made the tool available to about 10,000 scientists. The lab said last year that the technology was supposed to help accelerate its research efforts “in the domains of nuclear deterrence, energy security, materials science,” and other areas.
Anthropic has also worked with the National Nuclear Security Agency on evaluating potential AI-related nuclear safety risks. For example, the agency has provided Anthropic with “high-level” metrics and guidance that have helped the company analyze the threat of its own technology. Anthropic has also worked with the NNSA on developing technology that can scan and categorize AI chatbot conversations and search for signs that someone might be using an LLM to discuss building a nuclear weapon.
