Sacks Calls for 'Bretton Woods of Super Intelligence'
Synopsis
Key Takeaways
The man shaping Washington's artificial intelligence strategy just dropped a four-word phrase that could define the next decade of global tech governance. White House AI and Crypto Czar David Sacks posted 'The Bretton Woods of Super Intelligence' on X on Wednesday, September 30, 2026 — a pointed invocation of the most consequential international economic summit in modern history, applied now to the existential question of superintelligent AI.
Why 'Bretton Woods' Is the Sharpest Possible Frame
In July 1944, representatives of 44 Allied nations gathered in Bretton Woods, New Hampshire, and in three weeks built the scaffolding of the post-war world economy — the International Monetary Fund, the World Bank, and a dollar-anchored exchange-rate system. It was coordination under pressure, fast and consequential, before the world fractured again. Sacks is suggesting superintelligence demands exactly that kind of moment: a convening of nations, a negotiated framework, institutions with teeth.
The comparison is not casual flattery toward history. Bretton Woods worked because the stakes were understood to be civilisational, because a dominant power (the United States) was willing to set the table, and because the alternative — fragmented, competing monetary regimes — was clearly worse. Sacks's framing implies all three conditions may now apply to AI.
The Governance Gap Superintelligence Is Racing Into
Multilateral AI talks have been underway for several years — safety summits, standards bodies, voluntary commitments from frontier labs — but no binding international architecture exists for superintelligence specifically. The gap between the pace of AI capability and the pace of governance is widening. A Bretton Woods-style framework would, in theory, set rules on development timelines, safety thresholds, compute access, and liability before a system crosses the threshold into superintelligence — not after.
As the Trump administration's point man on both AI and crypto, Sacks sits at the intersection of Silicon Valley's frontier labs and the executive branch's foreign-policy machinery. A post of this weight from that role is not idle musing — it signals a direction of travel for U.S. policy on global AI coordination.
What a 'Super Intelligence Summit' Would Actually Mean
Translating the metaphor into mechanics is where the hard questions live. Bretton Woods succeeded partly because monetary policy, however complex, was a bounded domain. Superintelligence governance would need to cover compute supply chains, model weights, alignment research standards, military applications, and the rights of nations without frontier AI capabilities. India, for instance, has consistently argued in multilateral forums for an equitable AI order that does not entrench a two-tier world of AI haves and have-nots — a dynamic a Bretton Woods-style process could either resolve or calcify.
Sacks's post does not specify a venue, a timeline, or a list of nations. What it does is put a name to an ambition — and in policy, naming something is often the first move toward building it.
If 1944's Bretton Woods prevented a fragmented global economy, the pressure is now on its 21st-century equivalent to prevent a far more fragmented — and far more dangerous — race to superintelligence.