Polymarket US Files AI Compute Contracts as CFTC Seeks Standards for Emerging ‘Digital Oil’ Market
Polymarket US has self-certified a brand new class of contracts tied to the long run worth of synthetic intelligence computing energy, becoming a member of a rising race to construct monetary markets round one of many AI trade’s most vital and costly sources.
The filing comes as the Commodity Futures Trading Commission (CFTC) begins creating a regulatory framework for the rising market. In a Bloomberg Television interview Thursday, CFTC Chair Michael Selig known as compute “an important commodity, I believe, of our time” and “basically a digital oil.”
“We need the costs to be found for this worthwhile commodity within the United States, the benchmarks to be right here,” Selig stated. “We’ve gotta have these markets right here.”
Polymarket joins Kalshi, CME, ICE, Architect and potential change Liquid Compute in pursuing other ways to commerce or set up future costs for AI compute. The push comes as the CFTC’s new request for public comment raises questions on whether or not right now’s fragmented and sometimes opaque compute market is sufficiently liquid, standardized and proof against manipulation to help a mature derivatives market.
What AI compute markets are literally buying and selling
AI compute refers back to the processing capability used to coach and run synthetic intelligence fashions. Much of the highest-value capability right now comes from GPUs, or graphics processing models, high-powered chips able to performing enormous numbers of calculations concurrently. Nvidia‘s H100, H200 and newer B200 GPUs are among the many chips mostly utilized in large-scale AI infrastructure.
The rising market is usually not about buying and selling possession of the chips themselves. AI corporations, cloud suppliers and data-center operators purchase or lease entry to computing capability, typically priced by the GPU-hour, and the derivatives now being developed are designed to place a market worth on what that entry could price sooner or later.
That might permit an AI firm anticipating to wish massive quantities of GPU capability months from now to hedge in opposition to rising rental prices, whereas an infrastructure supplier might handle the danger that future costs fall. The CFTC stated in its request for remark that compute futures might support threat administration and worth discovery and permit monetary markets to “combination and reveal details about the way forward for the AI financial system.”
Different exchanges are approaching that job in several methods. Conventional futures can observe an index of GPU rental costs over time, whereas prediction market contracts can ask whether or not the value of a specific kind of compute can be above or under a specified degree on a future date. A sequence of contracts throughout completely different expiration dates can be used to assemble a ahead curve, displaying the place merchants collectively anticipate compute costs to be weeks or months forward.
Selig desires US to set the compute benchmark
Selig put compute alongside prediction markets, crypto and different rising merchandise in his remarks on the CFTC’s Innovation Advisory Committee meeting Thursday, framing the initiative as “Winning the AI Race: Roadmap for Compute Market Dominance.” He tied the hassle to the White House’s AI Action Plan, which calls for enhancing the monetary market for compute to broaden entry to large-scale computing sources.
The CFTC is working with the Department of Commerce on that effort and plans to make use of suggestions from its new request for remark to develop what Selig described as a “gold customary regulatory framework” for compute markets.
“Just as American markets helped set up the gold customary for buying and selling the commodities that powered the economic financial system, we’ll do the identical for the commodity that can energy the intelligence financial system,” Selig stated in a separate statement Wednesday.
His Bloomberg feedback a day later made clear that the target extends past merely permitting exchanges to record new merchandise. Selig stated he desires the underlying worth discovery and benchmarks established domestically, with compute buying and selling on clear U.S. markets and open order books.
“We’re working very arduous along with Secretary (Howard) Lutnick and the Department of Commerce to set the usual right here within the U.S.,” Selig stated. “We need the very best of the very best, and we’re gonna lead in these markets.”
CFTC sees hurdles in constructing the market
For all of Selig’s enthusiasm, the CFTC’s request for comment makes clear that regulators nonetheless see main challenges in turning compute right into a mature derivatives market. The company stated compute pricing stays fragmented and is commonly set via “opaque bilateral transactions,” limiting the quantity of public worth information out there to exchanges and regulators.
The Commission additionally questioned whether or not compute but has the fungibility, standardization and liquidity usually related to commodities underlying futures markets. Prices can range extensively relying on the GPU mannequin, supplier, area and contract phrases, whereas dominant suppliers could have sufficient pricing energy to affect the money market or a benchmark derived from it.
One of the RFC’s sharpest questions asks whether or not buying and selling ought to be permitted in a spinoff whose settlement worth depends on information the CFTC “could not have the ability to observe, confirm, or surveil.” The company can also be asking what protections would stop a compute supplier from influencing an index by altering posted charges, directing capability to or away from a venue or selecting whether or not to execute transactions throughout a settlement window.
The CFTC is searching for enter on these points together with buyer protections, market surveillance and perpetual compute futures. Comments are due Oct. 20, after which Selig has stated the company will proceed creating its regulatory framework for the market.
Prediction markets transfer into compute pricing
Polymarket US’s new filings cowl binary contracts tied to future values of AI compute indexes. One specifically certified contract asks whether or not the Ornn Data H100 SXM GPU Price Index can be a minimum of $2.50 per GPU-hour on March 15, 2027. A wider class certification would permit Polymarket US to record related contracts tied to H100, H200, A100, RTX 5090 and B200 GPU indexes. The broader submitting additionally covers Ornn indexes measuring the value of AI-model utilization by the token.
The contracts can ask whether or not an index can be above, under, a minimum of, at most, between or precisely a specified worth at a future level. Polymarket stated the merchandise could also be listed starting Aug. 24, though it stays to be seen whether or not Polymarket will launch them instantly or whether or not the CFTC will intervene whereas its compute remark course of is underway. That chance shouldn’t be purely theoretical: in July, the CFTC stayed a CME self-certified 24/7 crude-oil contract whereas a associated company remark course of was underway.
Kalshi has been pursuing a associated technique. In July, the change stated it was utilizing weekly and month-to-month prediction markets extending as far as a yr out to assemble a compute ahead curve, giving merchants a view of the place GPU rental prices are anticipated to maneuver.
“We are utilizing prediction markets to construct the ahead curve, which can present the market a view of what compute prices can be sooner or later for completely different grades and time-frames of GPUs,” Kalshi Chief Risk Officer Udesh Jha informed Bloomberg.
Jha stated the curve might finally help further merchandise, together with futures and choices. “It’s a key enabler for a variety of subsequent hedging, threat administration and even speculative actions,” he stated.
Exchanges race to cost AI compute
Prediction markets are just one a part of the rising competitors. CME Group plans to launch H100 and B200 rental-index futures on Oct. 5, pending regulatory evaluate, utilizing benchmarks from Silicon Data. ICE has announced GPU compute futures primarily based on Ornn indexes and a separate partnership with NATIVX for energy-normalized compute futures. Architect Financial Technologies is building a U.S. change for futures and choices tied to GPU rental prices and different AI supply-chain inputs.
Another potential change has been constructed round that concept from the beginning. DeFi Rate reported in February on the startup then working as Pluto, whose pending designated contract market and clearinghouse purposes are publicly listed underneath the names PMEX Markets and PMEX Clearing. The firm has since rebranded as Liquid Compute, which describes itself as the “monetary layer for AI compute” and is presently inviting market contributors to request access whereas its U.S. change purposes stay pending.
“The goal of the change is to show compute right into a monetary asset similar to oil, gold, (or) different commodities,” CEO Ronit Jain informed DeFi Rate on the time. Liquid Compute has since highlighted an H100 OTC forward trade involving Wintermute and said this week that it’s “dwell and reserving swaps now.” The firm has not publicly recognized the entity or jurisdiction via which these transactions are being performed.
The race now’s to construct the clear, trusted pricing infrastructure wanted for a mature AI compute market. The CFTC desires that market centered within the U.S., however its new remark course of reveals regulators are nonetheless deciding what requirements these costs and merchandise might want to meet.
The put up Polymarket US Files AI Compute Contracts as CFTC Seeks Standards for Emerging ‘Digital Oil’ Market appeared first on DeFi Rate.
