Sriram Krishnan Leaves White House to Launch Independent AI Policy Institute

Sriram Krishnan, the most recognisable figure in the technology ecosystem around Trump, is stepping down from his position as artificial intelligence advisor at the White House. According to TechCrunch reporting, Krishnan is not disappearing from the scene: he plans to establish his own institution through which he will continue shaping AI policy in the Republican administration. The news broke on 6 June 2026.

Krishnan joined the Office of Science and Technology Policy (OSTP) in the early stages of Trump's second term, coming from venture capital where he was a general partner at a16z, with a distinctly pro-deregulation profile and a belief in private leadership in AI development. His time in the administration has coincided with the reversal of much of the regulatory framework established by Biden, including the dismantling of several guidelines on risk assessment and transparency in large-scale models.

What the revolving door means

A senior advisor leaving an executive position to establish their own institution is not uncommon in Washington. What is noteworthy here is the stated purpose: not to break with the agenda, but to continue it from outside. This makes Krishnan's future institution more akin to an operational think tank than a traditional academic centre.

These types of structures, well-funded and with direct access to decision-makers while avoiding federal bureaucracy, have proven to be effective channels for sustaining policy lines beyond individual mandates. In AI, where innovation cycles vastly outpace electoral cycles, this continuity can prove more influential than the official position itself.

For the technology sector, particularly companies working with advanced language models and automation tools, Krishnan's move signals that the administration's deregulatory approach does not depend exclusively on his presence in government. The architecture of influence is diversifying.

Why it matters beyond the United States

US AI policy is not a domestic matter. Decisions on evaluation standards, disclosure requirements, or accountability frameworks that are adopted or rejected in Washington have ripple effects on European regulators, frameworks like the EU's AI Act, and, by extension, how global companies design their products.

In recent months, we have seen the regulatory divergence between the United States and Europe widen. While the EU advances with AI Act implementation imposing concrete obligations on high-risk systems, the Trump administration has chosen a position of minimal federal intervention. The institution Krishnan intends to establish will likely act as a counterweight to voices advocating for stronger regulation, both within and outside the country.

For engineering teams developing integrations with models like those from Anthropic or deploying agents in enterprise environments, this political context is not background noise: it determines what audit, transparency, or certification requirements might be mandated in different markets over the coming years.

Who stands to gain

Companies with high exposure to US government contracts or those reliant on a light regulatory environment to scale rapidly are most interested in Krishnan's policy line having continuity. So too are venture capital funds with AI positions, a world from which Krishnan comes and where he will likely maintain close ties.

Conversely, those advocating for more robust risk assessment frameworks—safety researchers, some European lawmakers, civil society organisations—will see this institution as an additional actor complicating efforts toward international harmonisation.

---

From ClaudeWave, the takeaway is pragmatic: the movements of figures like Krishnan matter more for the networks they activate than for the positions they hold. We will need to see what funding and alliances he brings together before assessing the real scope of this new institution.