18 months. That's how long Sriram Krishnan served as senior policy adviser on artificial intelligence before announcing he will step down at the end of June. In a post on X he said he will take a break and then work with outside institutions to tackle "large challenges facing America on AI," and described the role as "the privilege of a lifetime," language TechCrunch quoted. Krishnan said he credited President Donald Trump’s leadership for putting the United States "in the AI race," and he signalled plans to found an external organisation to continue technical and policy work.

18 months of policy work under Krishnan coincided with a string of White House moves that reshaped the regulatory terrain in ways many in industry had sought.

From Silicon Valley product chief to White House adviser

Krishnan joined the White House on 20 January 2025, when the second Trump administration began. His résumé before government spanned leading product teams at Microsoft, Twitter, Yahoo, Facebook, and Snap, and included a recent partnership at venture firm Andreessen Horowitz. Those credentials were part of the rationale for his appointment as senior policy adviser on artificial intelligence, even though critics in parts of the president’s political orbit objected to his selection in 2024.

Inside the administration Krishnan became a central figure in shaping a pro-industry AI agenda. He worked on the American AI Action Plan and on a National AI Policy Framework that accompanied the president’s executive orders on artificial intelligence. The shift was notable for its emphasis on infrastructure and industry acceleration, with the American AI Action Plan placing higher priority on data centre construction and capacity than on sweeping new federal regulation.

Policy choices that favoured industry priorities

The White House under Krishnan pushed a series of executive actions that echoed the industry playbook. Executive orders advanced steps to limit states’ ability to impose their own AI rules, and an order on AI oversight was narrowed and delayed after industry pushback, according to reporting that summarised the executive actions Krishnan later flagged as public accomplishments. A separate executive order directed federal agencies to ask leading AI developers to voluntarily submit their most capable models for government cybersecurity testing before public release.

Administration officials also publicly discussed broader interventions, including proposals that the U.S. government might take equity stakes in major AI firms, a concept the president praised when talking about partnerships between the public and private sectors.

Mastheads

The cumulative effect was to prioritise capacity and cooperation with industry rather than an immediate, comprehensive federal regulatory regime.

Those outcomes reflect the read Krishnan and his allies offered of the policy challenge: the immediate bottleneck was compute and infrastructure, not only rules. He framed his next steps in his X post around similar themes, saying he planned to focus on energy, data centre capacity, and wider public access to AI benefits. He didn't provide a launch date for the new organisation he has discussed.

Krishnan also noted his working relationship with investor David Sacks. Sacks stepped down earlier this year as the administration’s AI and crypto lead and now serves as co-chair of the President’s Council of Advisors on Science and Technology. Several accounts have emphasised that Krishnan and Sacks coordinated closely inside the administration.

Reaction from industry and from political corners was mixed. Supporters within the administration described Krishnan’s contributions as vital, and a White House spokesman told reporters the White House looked forward to working with him in any future capacity.

Critics questioned his appointment from the moment it was announced, reflecting an earlier controversy over the selection in 2024. Yet multiple outlets reported that he plans to remain engaged with the administration’s AI agenda from outside government.

The Washington Post reported that Krishnan intends to found an external institution that could continue to influence Trump-era AI policy and supply technical support to the government, a model described in that reporting as combining policy analysis with engineering capacity. Such an organisation, if realised, would mirror a pattern of outside groups offering both policy advice and technical assistance to federal agencies, and would allow Krishnan to continue exerting influence without the constraints of government service.

In his X post Krishnan said he would take a break after leaving at the end of June, then pursue institution-building to address the challenges he has raised publicly. He also thanked colleagues and cited the Trump administration’s focus on putting the United States "in the AI race," a formulation TechCrunch quoted when reporting on his remarks.

Krishnan leaves at a moment when the White House has made clear choices about how to marshal federal resources for AI. Those choices have tilted toward infrastructure, voluntary collaboration with leading developers, and executive action to pre-empt a patchwork of state rules. For proponents of a rapid industrial build-out these are wins. For those who pressed for an early, far-reaching federal regulatory framework they're unfinished business.

Related Articles

He will depart the White House at the end of June 2026, and has said he will take a break before pursuing an external organisation focused on data-centre capacity, energy and wider public access to AI benefits.

This article was created with AI assistance.