Commerce Department orders Kalshi to remove AI compute futures market, citing national security

Commerce Department Blocks Kalshi's AI Compute Futures Market, Citing National Security

The US Commerce Department has ordered Kalshi, a prediction market platform, to shut down its AI compute price-tracking market, citing national security concerns. The move caught industry observers off guard—the market itself appeared benign compared to other prediction products.

Kalshi's product pulled together data from several underlying markets that allowed users to bet on the cost to rent NVIDIA chips, creating a forward-looking indicator of where AI compute costs were heading. The platform complied quietly, though many of the underlying markets remained open for trading.

Commerce also pressured the Commodities Futures Trading Commission to freeze approval of new compute futures contracts for 60 days, potentially delaying plans by CME, NYSE parent Intercontinental Exchange, and other operators to list two-sided betting parlors. That freeze could eliminate compute prediction market futures entirely if no new contracts get approved before the window closes.

The stated rationale remains murky. A Commerce Department spokesperson denied ever asking Kalshi to remove the market, calling reports of the order "false to some degree." Kalshi declined to comment. Neither side has articulated why compute futures posed a national security problem—a national security framing that contrasts sharply with how the regulator treats other prediction markets on flight delays or other sensitive topics.

One theory circulating among market participants: compute futures could be manipulated to artificially show a sharp drop in the cost of older chips, destabilizing AI stocks and debt markets. Because these markets are thinly traded, even modest trading volume could trigger outsized volatility. A short seller betting compute costs would plummet could trigger a selloff in AI equities, profit during the panic, then unwind the position.

The stakes are real. Compute costs have become one of the most important numbers in the US economy. Older chips serve as collateral for billions of dollars of borrowing by what the industry calls "Neo Clouds." At the same time, companies adopting AI face uncertainty about whether power and infrastructure shortages will send prices soaring. A futures market would have let both sides hedge those bets and given traders a venue to discover true prices.

Instead, the market has been shut down before it could mature—leaving the actual cost discovery mechanism that underpins AI infrastructure investment in the hands of bilateral deals and spot markets, not transparent, regulated futures contracts.