The head of the Trump administration's AI safety agency CAISI has resigned after three months in the role, leaving the body without a permanent director. Arvind Raman, director of the National Institute of Standards and Technology, will serve as acting director, a spokesperson confirmed.
How standards bodies work and why this matters
Government AI safety oversight is not primarily a policy function. It is a standards function. The mechanism works like this: an executive-level directive that says "AI should be safe" has no operational meaning until a body like CAISI translates it into specific evaluation criteria and procurement guidance that agencies and vendors can act on. That translation layer is slow by design. It requires extended comment periods and interagency coordination that each run over months.
The constraint is institutional continuity. A director who departs after three months hands off to a successor who must restart the relationship-building that standards work requires and re-establish the agency's credibility with external stakeholders. None of that is fatal, but it costs time measured in quarters.
Raman's interim appointment
Raman's role as NIST director already places him at the center of the federal government's technical standards infrastructure. Appointing him acting director of CAISI is an administratively logical bridge: it avoids leaving the agency without any leadership chain while a permanent selection is made.
The spokesperson provided no timeline for naming a permanent director. The source does not identify the outgoing director by name.