Sriram Krishnan is leaving his role as White House AI advisor

The more consequential detail isn't the departure itself but the institutional form Krishnan is choosing: a new independent organization rather than a return to a firm like a16z, where he previously worked. That structure is designed to carry policy influence without the disclosure obligations or conflict-of-interest optics that come with a direct industry affiliation.

This sits directly alongside OpenAI's June 1st formalization of its own regulatory engagement posture, covered here under 'Our views on AI policy and political advocacy.' Both moves reflect the same underlying dynamic: as federal AI policy becomes higher-stakes, the organizations best positioned to shape it are building dedicated infrastructure outside the normal lobbying or government-employment channels. Krishnan's new institution and OpenAI's advocacy clarification are parallel bets that independent, well-resourced policy shops will matter more than internal government roles as legislation accelerates. The Import AI digest from June 1st also flagged that AI oversight is operationally difficult, which gives Krishnan's exit a structural read: the White House role may simply lack the institutional capacity to do what he wants to do.

Watch whether Krishnan's new organization takes formal positions on the federal AI legislation expected in late 2026. If it does, and those positions align closely with frontier lab preferences, the 'independent' framing will face real scrutiny.