The Commodity Futures Trading Commission is asking for public feedback on a new type of contract tied to the cost of AI computing power. The Chicago Mercantile Exchange is planning to launch such futures in October. The review comes as high-performance computing capacity grows harder to secure.
Why the CFTC is weighing in
The agency's request for comment opens a formal review of how these contracts should be regulated. It's an early step, but it could shape whether AI compute futures become a standard tool for businesses that need to plan around rising and volatile computing costs.
The CFTC hasn't said exactly what it wants to know. But the move signals that regulators are paying attention to a market that barely exists yet. AI computing power is increasingly scarce, and that scarcity is driving interest in ways to hedge against price swings.
What CME is planning
CME has its eye on an October launch. That timeline puts the exchange ahead of most competitors, though the regulatory process could slow things down. The contracts would let buyers and sellers agree on a price for computing power at a future date, similar to how farmers lock in grain prices or airlines hedge fuel costs.
For companies building and running large AI models, the ability to lock in compute costs could be significant. Training a single large model can consume enormous amounts of processing power, and those costs have been climbing as demand outstrips supply.
The scarcity factor
The whole idea hinges on scarcity. If AI computing were abundant and cheap, there'd be little reason to hedge it. But the opposite is true. Chips, data centers, and the energy to run them are all in tight supply, and that's made compute costs unpredictable.
A futures market wouldn't make computing power easier to find. It would just give buyers and sellers a way to manage the financial risk. That's a different kind of value, but one that could matter just as much to the bottom line.
The CFTC's review will determine how these contracts fit into existing rules. The agency has to decide whether AI compute is closer to a traditional commodity like oil or something more abstract. That decision will affect everything from margin requirements to who's allowed to trade.
Public comments are the next step. The CFTC hasn't set a deadline yet, but once it does, market participants will have a chance to weigh in. CME's October target could slip if the review drags on.
