Updated
Updated · Computerworld · Jul 21
US AI Testing Institute Chief Resigns After 3 Months, NIST's Raman Takes Interim Role
Updated
Updated · Computerworld · Jul 21

US AI Testing Institute Chief Resigns After 3 Months, NIST's Raman Takes Interim Role

3 articles · Updated · Computerworld · Jul 21

Summary

  • Chris Fall resigned about three months after taking over the Center for AI Standards and Innovation, the US body that tests advanced AI models for safety and security.
  • Arvind Raman, the current NIST director, will serve as acting CAISI chief while still overseeing the Commerce Department office that houses the institute; Commerce gave no reason for Fall's departure.
  • CAISI, reorganized in April from the former US AI Safety Institute, works with Anthropic, Google DeepMind and OpenAI on voluntary evaluations covering cybersecurity, misuse, reliability and other frontier-model risks.
  • Sanchit Vir Gogia of Greyhound Research said the key issue for enterprises is whether CAISI keeps its methods, publications and research teams stable, because government evaluations are only one input into internal AI governance.
  • The leadership change lands as Commerce takes a more active role in advanced-AI policy and national-security reviews, making continuity at the testing institute more important than the individual running it.

Insights

With its revolving door of leaders, can the top US AI safety institute still be trusted to evaluate frontier models?
Amid leadership turmoil, is America losing ground in the global race to govern and secure powerful artificial intelligence systems?