A 19-page request for comment issued by the Commodity Futures Trading Commission in Washington, DC on August 19, 2026 by the Commission, over the issuance line of Christopher Kirkpatrick, Secretary of the Commission. It carries RIN 3038-AF77, touches 17 CFR parts 1 and 38, and was published in the Federal Register on August 21, 2026 at 91 FR 54259 as FR document 2026-17163, with comments due 60 days after publication. The comment docket is CFTC-2026-1850 (Source: cftc.gov).
The document is a request for comment, not a proposed rule. It proposes no regulation, approves no contract, and takes no position on the listing of any particular product. Its stated purpose is "to better inform [the Commission's] understanding and oversight of derivatives markets in compute," and it invites comment "on the specific questions set forth in Section II … as well as on all aspects of the compute markets (including those not mentioned herein)."
Framing
The request opens on the Commission's statutory mandate: the derivatives markets it oversees are "affected with a national public interest" because they facilitate risk management and price discovery "through trading in liquid, fair and financially secure trading facilities" (CEA section 3, 7 U.S.C. 5), and section 3(b) charges the Commission with deterring manipulation, ensuring transaction integrity, avoiding systemic risk, and promoting responsible innovation and fair competition.
It then defines its subject as "an emerging class of derivatives contracts that reference the price of access to computing power ('compute'), which is the processing power primarily used by the large language models … at the center of the artificial intelligence … economy," and states that because AI "anchors a large and growing share of the U.S. economy and public equity markets," compute "has become a multi-hundred-billion-dollar enterprise and is a scarce, capital-intensive commodity." A footnote attaches a figure to that scale, citing Federico M. Bandi, (Early) AI Compute Asset Pricing (July 1, 2026): the 2025-Q4 installed compute stock implies "a gross compute service flow of around $430 billion to $1.3 trillion per year, or approximately 1.4% to 4.0% of U.S. GDP."
The policy context invoked is the July 2025 White House AI Action Plan — quoted for "whoever has the largest AI ecosystem will set global AI standards and reap broad economic and military benefits," and cited for its recommendation to ensure startup and academic access to large-scale compute by improving the financial market for compute — together with Executive Order 14179. The Commission characterizes these as echoing the purposes Congress codified in the Act in 1974. The document was designated a significant regulatory action under section 3(f) of Executive Order 12866 and reviewed by the Office of Management and Budget.
The designated-contract-market framework recited
Roughly half the document restates the existing regime. Designated contract markets (DCMs) are CFTC-regulated exchanges that must comply with 23 statutory Core Principles under CEA section 5(d) and the implementing rules at 17 CFR part 38. They are self-regulatory organizations, which the request underlines by quoting a 2010 Commission notice: each DCM "is not simply a corporation, but a corporation charged with the public trust."
The provisions the request treats as load-bearing for a novel underlier are Core Principle 2 (establish, monitor and enforce compliance with exchange rules), Core Principle 3 (list only contracts not readily susceptible to manipulation), Core Principle 4 (capacity and responsibility to prevent manipulation, price distortion, and disruption of the delivery or cash-settlement process), Core Principle 5 (deliverable supply and position limits), Core Principle 9 (a competitive, open and efficient market protecting price discovery), and Core Principle 11 (financial integrity of transactions through a derivatives clearing organization). A DCM may list a new contract by self-certifying compliance under 17 CFR 40.2 or by seeking Commission approval under 17 CFR 40.3; either way the terms and conditions must be filed before listing.
For cash-settled contracts the request leans on the Appendix C guidance to part 38, which asks that the settlement price specification fully describe the essential economic characteristics of the underlying commodity and how the final price is calculated, and which treats a cash-settled contract's utility for risk management and price discovery as "significantly impaired if the cash settlement price is not a reliable or robust indicator of the value of the underlying commodity." The Commission states its preliminary understanding that most compute derivatives would initially be cash-settled, because of "the potential infrastructure-related challenges associated with delivering the underlying commodity."
Preliminary findings on compute markets
The substantive core is a short section of preliminary beliefs, each hedged as preliminary.
On the case for such a market: because the price of compute "is one of the most material costs of AI that affects the entire AI production stack," a compute futures market may provide a means of managing price risk, discovering prices, and disseminating pricing information about general trends in AI adoption. Quoting Bandi, the request says compute futures would allow financial markets "to aggregate and reveal information about the future of the AI economy." A footnote reproduces Bandi's placement of the compute market "at the interface between infrastructure and model production," so that its price carries information about the cost of deploying and operating data centers on the supply side and the productivity of AI applications on the demand side, and attributes to Nvidia chief executive Jensen Huang the description of the AI stack as a five-layer cake of energy, chips, infrastructure, models, and applications.
Against it, the Commission sets out four difficulties: compute markets are fragmented; price formation "primarily occurs in opaque bilateral transactions," which limits current and historical price data; consensus is "still forming concerning the appropriate underlying 'compute' commodity"; and dominant participants "may wield significant pricing power that may lead to manipulability, preferential pricing arrangements, and, in turn, unfair market dynamics." Pricing also "can vary dramatically across providers, regions, and contract structures."
From these the Commission draws the finding the document is most likely to be cited for: that "compute may not yet exhibit certain of the characteristics of commodities that typically underlie a commodity derivatives market, including fungibility, standardization, and sufficient liquidity." It treats the issues as "novel in some respects, but also informed by historical antecedents," citing Dennis W. Carlton's 1984 history of futures markets, the account of 19th-century exchange standardization in Merrill Lynch v. Curran, 456 U.S. 353 (1982), and a 2012 Commission proposal on not-for-profit electric utilities for the proposition that electric energy, like compute, cannot be bought in commercial quantities ahead of time and stored.
On what the underlier would be, the Commission's preliminary understanding is that a compute futures contract would typically reference "access to rented compute capacity from hardware the purchaser of such capacity does not own (e.g., the hourly rental price of compute from a B200)," while noting that the underlier "may be a different type of compute-related commodity, such as access to a stated volume of LLM inference tokens." Since the price would likely be "derived from a bundle of factors (e.g., provider, region, contract structures)," the Commission's preliminary belief is that a compliant market "would likely require standardizing variables typically associated with compute," both in the settlement index and in any physical-delivery specification.
The questions
Section II poses questions in four groups.
Compute cash markets: size, liquidity, and other considerations (seven questions). These ask how observed price behavior compares with cash markets the Commission customarily regulates, distinguishing on-demand, spot, reserved and committed purchase modes by volume, counterparty count, and whether any transaction price is published; what data sources should determine market size, liquidity, participants and supplier concentration, distinguishing public data from data derived from non-public bilateral agreements that in the Commission's preliminary understanding "carry the majority of economic value but tend to be undisclosed and negotiated privately"; what proportion of transactions occur at publicly disclosed prices, and "would it be appropriate to permit trading in a derivative contract settling to a price computed from data that the Commission may not be able to observe, verify, or surveil, in whole or in part"; whether any audits or independent verifications of that transaction data exist; how compute cash markets differ on storability, observable transaction record, producer concentration, cross-producer fungibility, and the existence of any price reporting agency or enforced standardized methodology; what effects listing would have back on the cash market, including whether it would change provider incentives on posted rates, bilateral reservation structuring, disclosure of utilization and committed capacity, or capacity allocation, and how to consider that "the parties best positioned to influence the reference price are the same parties that supply capacity"; and what deliverable-supply estimation methodology would support position limits under Core Principle 5.
Market oversight and susceptibility to manipulation (nine questions). These ask what a contract cash-settling to an index computed over predominantly bilateral, privately priced transactions must demonstrate to satisfy Core Principle 3; what would prevent a capacity provider from manipulating a settlement index "by adjusting a posted rate, directing capacity onto or away from a venue whose transactions the index calculation methodology treats as input data, or by executing or declining to execute transactions during the observation window," given that published compute price series are built in whole or part from rates the providers themselves administer with the remainder executed on venues a small number of participants operate or dominate; what volume, transaction-frequency and contributor-concentration thresholds would be appropriate; what surveillance capabilities Core Principle 4 would require and whether they are "presently feasible from a technological, operational, and legal perspective," including whether a DCM should be required to maintain information-sharing arrangements with each compute venue and capacity provider feeding the settlement price; what recalibration of safeguards and liquidity protections would keep thin liquidity from producing "disproportionate or runaway price movements"; what trader or intermediary conduct patterns in compute cash markets warrant close monitoring; whether any existing cash price series meets the Appendix C reliability, acceptability, public-availability and timeliness criteria, and "what steps should the Commission take, if any, if no such cash price series is available"; what the estimated deliverable supply for compute is and how it is measured; and whether any other Core Principle considerations apply.
Customer protection (five questions). These cover heightened anti-money-laundering and know-your-customer concerns relative to mature commodity markets and the BSA/AML burden on introducing brokers and futures commission merchants; disclosure and other customer-protection considerations arising from a contract "settling against a geopolitically sensitive commodity, including to retail participants," and how these differ from contracts on commodities such as oil; whether contract terms should be required to disclose idiosyncratic risks; what protections are appropriate for retail users; and what conduct particular to compute markets might require further Commission action.
Perpetual compute futures (two questions). Whether perpetual contracts would offer advantages over fixed-date futures and meet risk-management needs existing products cannot, and whether they pose unique risks warranting additional safeguards.
Section III adds a general request for comment on all aspects of the questions, particularly encouraging "empirical and data-driven input."
Provenance
The Commission's own copy of the document as approved is at cftc.gov and carries the running header "As approved by the Commission (subject to Office of the Federal Register technical corrections)." The Federal Register version is the official publication. The voting summary reported in press coverage — that Chairman Michael S. Selig voted in the affirmative with no commissioner voting in the negative — appears in the Federal Register presentation, not in the CFTC copy, which carries only the Secretary's issuance line. Chairman Selig's statement that "America cannot win the AI race without a robust derivatives market for compute" comes from CFTC Release 9286-26 of the same date, not from the request itself (Source: cftc.gov).
The request does not address the CME Group–Silicon Data compute futures announced for October 5, 2026 pending regulatory review, which are a separate matter; the 60-day comment period running from August 21, 2026 closes after that announced launch date (Compute derivatives).
Relationships
- supports: Compute derivatives — supplies the regulatory framework and the Commission's preliminary findings
- depends-on: America's AI Action Plan, Executive Order 14179 — Removing Barriers to American Leadership in Artificial Intelligence — cited as the policy basis for Commission action
- related: Commodity Futures Trading Commission (CFTC), Compute Governance, Inference Economics and Token Pricing, Kalshi