Anthropic is closely tracking how Claude and AI are changing the economy. Now, the leading AI company wants to know what’s in store for democracy.
Anthropic, which confidentially filed an S-1 draft as it moves toward a blockbuster IPO, is hiring for a newly created team focused on “AI and the rule of law.”
“@AnthropicAI has studied what AI means for the economy. This team asks a different question: what will it mean for executive power, for courts and elections — and for the public deliberation that constitutional democracy ultimately rests on?” Matthew Botvinick, a Resident Fellow at Yale Law School who will lead the team, wrote on X.
According to the job posting for a staff position, Anthropic will pay between $295,000 to $345,000. Candidates should have a law degree, or a degree in political science or a similar field at the Ph.D. level or equivalent, or “extensive government experience at a leadership level.”
“The ideal candidate understands the technical landscape well enough to reason about AI capabilities and risks, and understands democratic institutions well enough to see where those risks become structural threats,” the posting reads.
Anthropic laid out four areas it wants “the AI & Rule of Law” team to focus on: “AI safety evaluations with a legal alignment lens, institutional vulnerability analysis, novel legal issues in frontier AI, and applications that bolster democratic processes.”
Some of those issues are already playing out.
The extent to which AI systems adhere to “constitutional norms” and, more importantly, how AI companies train their models on those norms, were part of the disagreement between Anthropic and the Pentagon that led to the Defense Department effectively blacklisting the company. Anthropic has challenged that decision, and its litigation remains ongoing. At the time, Pentagon R&D chief Emil Michael said he was concerned about how Anthropic trains Claude based on its internal constitution.
As for novel legal issues, in March, Nippon Life Insurance sued OpenAI in federal court in Illinois, seeking to hold the AI company liable for a woman who repeatedly used ChatGPT to draft legal documents to try to reopen a settlement agreement, even though the insurance company said the agreement could not be altered. In May, OpenAI moved to dismiss the case, stating in part, “ChatGPT is not a lawyer, and it does not practice law.”
