Krishnan Exit Shows the Revolving Door Inside AI Policy
Reuters reported Saturday that Sriram Krishnan, a White House adviser focused on artificial intelligence policy, is leaving his position. Washington Post and CNBC coverage framed the departure as the exit of a high-profile Trump technology adviser after a short but consequential period shaping federal AI direction. The story lands as Washington is increasingly treating AI not as a narrow software issue but as industrial policy: compute access, export rules, defense use, data-center power, procurement, and public-private partnerships are all being pulled into the same policy lane. The personnel change matters because the next phase of AI governance will be decided less by abstract principles than by who has the ear of the administration, which companies get treated as national champions, and how quickly federal agencies turn policy memos into procurement and regulatory decisions. For citizens, the important question is not whether one adviser stays or goes. It is whether AI policy becomes a public-interest framework or a rotating-door contest among firms trying to become the operating system of government.
The public tends to read a personnel story as inside baseball. One adviser leaves, another adviser arrives, and the headline disappears by Monday. That is the wrong way to read AI policy now. The real story is that the center of gravity has moved. Artificial intelligence is no longer a Silicon Valley product category waiting for Washington to notice it. It is becoming a governing layer, and the people who sit between companies and the state are becoming some of the most important policy actors in the country.
That is why the Krishnan departure is worth more than a quick nod. The White House has been moving toward a model where AI is tied to national competitiveness, defense, industrial strategy, energy demand, and public-private partnership. In that world, personnel is policy. Whoever helps define the rules also helps define which companies get access, which risks get downplayed, which competitors get fenced out, and which agencies become dependent on private systems they barely understand.
The incentive problem is obvious. The faster AI becomes strategic infrastructure, the more every major company wants to be treated as indispensable. Nobody wants to be regulated like a risky vendor. Everyone wants to be recognized as a national asset. That creates a subtle form of capture that does not always look like old-fashioned lobbying. It looks like advisory boards, fellowships, policy shops, emergency memos, procurement pilots, and revolving-door expertise that makes private interests appear indistinguishable from public capacity.
Citizens should care because the bill will not arrive as one big AI law. It will arrive in smaller decisions: which models agencies buy, what data those models touch, how much power data centers can pull from strained grids, whether national-security exemptions become a loophole for ordinary accountability, and whether state governments are allowed to set their own rules. Each choice sounds technical. Together they decide whether democratic oversight keeps up with the systems being embedded into public life.
The easiest mistake is to turn this into a personality story. The harder and more useful question is what kind of institutional structure is being built. If every important AI role becomes a short stop between venture capital, Big Tech, and federal power, the public will get policy by network effect. The firms closest to the room will define the public interest in language that just happens to match their business model.
A serious AI policy would start from the opposite premise. It would assume that scale creates dependency, dependency creates leverage, and leverage requires transparency before the public is locked in. The question is not whether America should lead in AI. The question is whether leadership means building accountable infrastructure or simply deputizing the largest platforms to govern faster than Congress can understand. Krishnan leaving the White House is a personnel move. The revolving door around AI policy is the institution citizens should be watching.