White House AI adviser Sriram Krishnan exits after 18-month stint

Krishnan’s departure matters less for day‑to‑day White House staffing gossip and more for what it signals about the maturation of AI policy as a specialized power center. As Senior AI Policy Advisor, he helped architect the American AI Action Plan and the National AI Policy Framework executive order, and served as a bridge between Silicon Valley, energy planners and national security hawks.([ndtv.com](https://www.ndtv.com/world-news/indian-american-sriram-krishnan-quits-as-white-house-ai-adviser-after-18-months-11601858?pfrom=home-ndtv_topstories_lastestImg&utm_source=openai)) Losing that blend of technical fluency and political trust at precisely the moment the administration is contemplating equity stakes in frontier labs creates a temporary vacuum.

For the race to AGI, this is a reminder that institutional capacity—not just model weights and GPU counts—will shape who can set norms for powerful systems. Krishnan has been a public advocate for an “American AI stack” spanning chips, model providers and cloud infrastructure; taking that worldview outside government likely means a new think tank, fund or industrial policy vehicle focused on making US‑aligned AI more coherent. In the short term, though, the departure could slow follow‑through on complex initiatives like pre‑deployment vetting of top models, international AI summits and energy policy tailored for data center build‑out.

The deeper question is whether this role becomes institutionalized—like the National Security Advisor—or remains a personal fiefdom that rises and falls with individual technocrats. If the latter, policy continuity on AGI‑relevant issues like safety standards, compute controls and public‑wealth mechanisms will be more fragile than many assume.