Bridgewater’s Greg Jensen: treat AI labs like systemically important banks if they own >5% of compute
What's the deal? Bridgewater managing CIO Greg Jensen argues that any firm controlling more than ~5% of U.S. or global AI compute should face heightened oversight “like we do with systemically important banks,” and that caps on compute ownership may be needed — in a The Information Q&A with Dakin Campbell.
- Concentration warning: Jensen projects OpenAI and Anthropic could control 35–50% of world compute within ~two years; argues a bipartisan case against monopolistic compute control.
- Open-source RL risk: Bridgewater finds private RL on open models “extremely effective” and currently hard to police — “slowing frontier labs alone is not nearly enough.”
- Bridgewater AIA Labs (founded 2023): 80+ employees; dedicated AI fund managing ~$4.5B; investment decisions in that stream are machine-made, with risk controls, data acquisition and trade execution kept human-driven (plus kill switches).
- Compute posture: Bridgewater rents compute rather than owning GPUs/DCs; partners include Thinking Machines Lab for RL tooling; constantly benchmarks Anthropic vs OpenAI rather than locking in.
- Portfolio view: AI infrastructure build-out trade is largely priced in; firm holds a “very small position” and prefers disruption/adoption trades; models DC build through 2028–29 and flags funding/execution/regulation — not a bubble call — as key risks.
Caveats: Interview/opinion. The 5% threshold and 35–50% concentration figures are Jensen’s estimates, not regulators’ rules. How a compute “SIFI” test would be measured (training vs inference; owned vs reserved) is unspecified. China-cooperation comments in the piece are speculative.
The signal: A top macro allocator is importing bank-style systemic-risk language into frontier AI — and Bridgewater’s own AIA stack (rent-vs-own, multi-model RL, Thinking Machines) shows how large allocators actually consume frontier compute.
Source: The Information — Bridgewater’s Greg Jensen on regulating AI firms like SIFIs.
