Anthropic has hired a former state Supreme Court justice as its first director of global affairs, a role created for the occasion. A hire doesn't normally make the news. This one says a lot about where AI labs now see their main risks.
Two years ago, the hires that got people talking in this sector were researchers. Today, they're lawyers and diplomats. That shift is worth pausing over.
What the role covers
A global affairs role typically spans three areas: relations with regulators, legal strategy, and dealings with foreign governments.
Creating this position and putting a senior former judge in it signals that these three issues have become priorities, and that they're being treated as a coherent whole rather than as separate support functions.
Why now
The context largely explains this decision, and we've documented it throughout the summer.
The conflict with the state. Anthropic found itself in open litigation with the Pentagon over a company's right to set usage limits on its technology, then suffered a global suspension of two models on a government decision. These aren't commercial disputes; they're matters of public law.
Regulatory opacity. As we've noted, the evaluation framework for frontier models was finalised without its contents being published. In an environment where the rules aren't public, the ability to talk to those who write them becomes a strategic asset.
International fragmentation. Europe applies its regulation, China builds its own framework, each jurisdiction develops its requirements. A company operating globally now has to navigate between divergent regimes.
There are two readings, and they aren't mutually exclusive. The favourable reading: a company that takes its legal framework seriously is giving itself the skills to do it properly, which is better than discovering the rules after breaking them. The more critical reading: recruiting at the top of the judiciary is also a way of gaining the means to influence the rule rather than simply comply with it. Both dynamics exist in every regulated industry, and AI won't be an exception.
The shift you need to see
This hire fits into a broader movement we've been observing for weeks. Competition between labs is no longer just about model capabilities, where gaps are narrowing and prices are collapsing.
It's now about access to regulated markets, the ability to secure approvals, mastery of legal risk, and the relationship with states. That's exactly what we described regarding compute sovereignty: the attributes of power are no longer technical.
For an observer, this means tracking appointments and regulatory filings as closely as model releases. What gets decided in an office in Washington or Brussels will likely weigh more heavily, over the next three years, than three points gained on a benchmark.