In the same week, the CFTC asked the public to help it write rules for a derivatives market that barely exists, and proposed cutting registration requirements for a category of market participant that has existed for decades. Chairman Michael Selig calls both moves part of the same project. I do not think they are, and the difference matters for anyone trying to guess where this regulator is actually headed.
The expansion side
On August 19, the CFTC published a formal request for comment on listing compute derivatives contracts, an early step toward building a rulebook for a market in raw AI compute capacity that does not meaningfully exist yet. “America cannot win the AI race without a robust derivatives market for compute,” Selig said in the agency’s announcement, framing the request as the first step toward “clear rules of the road” for trading compute the way markets trade oil or wheat.
That request lands on top of a pattern the agency has been building for weeks. It has already drawn a federal line around prediction markets, used emergency authority to intervene in one, and then tightened self-certification filing rules for the incentive programs prediction platforms use to attract traders. Our own cross-press look at that stretch of activity concluded that prediction markets got rules before they got banking access. Compute derivatives is the same playbook applied to a market that is even younger: build oversight machinery ahead of the product, not after a blowup forces the agency’s hand.
Advertisement
The contraction side, and the counter-argument
One day earlier, on August 18, the same chairman proposed the opposite move in a different corner of the agency’s remit: exempting SEC-registered investment advisers from also having to register as commodity pool operators, plus a matching exemption for commodity trading advisors, plus an inflation adjustment to the threshold for the existing small-pool exemption. “This proposal is yet another step to unwind overregulation and cut red tape for American businesses while still preserving market integrity,” Selig said of that plan.
The obvious reading is that these two moves cancel out, or that the deregulation is the real signal and the new-market oversight is just process, comment periods that may or may not produce binding rules. I think that reading gets the balance backward, and it is worth taking seriously before dismissing it, because the CPO and CTA exemptions are genuinely duplicative. An adviser already registered with the SEC and subject to its examination regime gains little from also filing a separate CFTC registration for the same pool. Cutting that is a defensible efficiency gain, not a retreat from oversight of anything that was actually being caught.
Why the expansion is the bigger move
Weigh what each side actually creates. The CPO and CTA change removes a filing requirement for pools that were already supervised by the SEC. The compute derivatives request opens the door to an entirely new contract category, one with no trading history, no established manipulation patterns and no settled answer on what counts as delivery for a commodity that is really just data-center time. Building that rulebook from a comment period, rather than after a market already exists and has already had a problem, is a much bigger undertaking than trimming a duplicate filing form. The agency is not shrinking. It is re-pointing its finite oversight bandwidth away from a compliance requirement that had stopped doing useful work and toward the categories, prediction markets and compute, where it clearly expects volume to show up next.
There is a version of this argument that says none of it matters yet, since a request for comment is not a rule and could sit for years without producing one. That is a fair caution, but it undersells how the CFTC has actually behaved this month. This is not an agency that floats ideas and waits. The same chairman used emergency authority against a prediction market with almost no notice, then followed with a self-certification advisory within days. A comment period from this CFTC has tended to be the opening move in a sequence, not a placeholder.
What it means for the finance leader
Firms trading either prediction contracts or anything compute-adjacent should treat “cut red tape” language from this CFTC as a signal about priorities, not a signal about the direction of the agency’s overall reach. Selig’s CFTC is telling the market plainly where it intends to spend its attention over the next year, and it is not on the paperwork it is removing. Compliance teams tracking this chairman’s record should weight the RFCs and self-certification advisories more heavily than the exemption proposals when forecasting what next year’s rulebook will actually cover.
Source: CFTC
