GENIUS Act: What Your Exchange Must Prove to List USDT in 2027

The date written in the GENIUS Act rule is July 18, 2028, but the provision affecting USDT carries no date: it applies starting January 18, 2027. On August 18, 2026, the Department of the Treasury proposed a new part 1523 of the Code of Federal Regulations to implement Section 3 of the GENIUS Act, featuring two distinct prohibitions on digital asset service providers (exchanges, custodians, payment gateways). § 1523.3(a) literally states "Beginning on July 18, 2028" and requires that the stablecoin be issued by a permitted issuer. § 1523.3(b) carries no date: it prohibits offering or selling a foreign-issued stablecoin in the U.S. unless that issuer has the technical capacity to comply with legal orders, and the preamble clarifies that this second prohibition "unlike section 3(b)(1) becomes applicable on the effective date of the Act"—January 18, 2027, according to note 10 of the document itself. An eighteen-month difference, with $183.447 billion of USDT on one side of the line and $184 million of USAT on the other. What follows is what an exchange will have to prove in January 2027, what due diligence § 1523.3(c) requires, and what remains excluded under § 1523.4(c)(3).

Why does the GENIUS Act clock biting USDT start on January 18, 2027?

Section 3(b) of the GENIUS Act contains two prohibitions that are often read as one; the proposed text separates them into two consecutive paragraphs with different temporal treatments.

§ 1523.3(a) is the general rule: starting July 18, 2028—"the date that is three years after the date of enactment," the preamble states—a provider may not offer or sell a payment stablecoin to a person located in the U.S. unless it was issued by a permitted payment stablecoin issuer (an issuer authorized under the law) or a foreign issuer meeting the criteria of section 18(a).

§ 1523.3(b) is the specific rule for foreign issuers, and it begins with "Except in accordance with § 1523.4." No date. The Treasury preamble leaves no room for interpretation: section 3(b)(2), "unlike section 3(b)(1)," becomes applicable on the effective date of the Act. And note 10 sets that date: "The effective date of the GENIUS Act is expected to be January 18, 2027 (i.e., the date that is 18 months after the date of enactment)," citing section 20 of Pub. L. 119-27. The "expected" has its reason: that section 20 sets the effective date as the earlier of two dates—18 months after enactment, or 120 days after primary regulators issue their first final rules—meaning January 18, 2027, can only be moved forward, never delayed.

The practical order is inverted compared to what the most cited calendar suggests. The first thing to arrive is a filter on the foreign issuer, a year and a half before the permitted issuer requirement comes into play in July 2028. It is the direct continuation of the case we followed on July 18, 2026, when seven agencies reached the law's anniversary without a single final rule: this is the Treasury's first regulatory text and the first to develop section 3 of the law, though not the first in the GENIUS case file—the Office of the Comptroller of the Currency (OCC) had already proposed its own on March 2, 2026, regarding outstanding issuance value.

What must an exchange prove to continue listing a foreign stablecoin?

§ 1523.3(b) requires the provider to ensure the foreign issuer "has the technological capacity to comply, and will comply, with the terms of any legal order and any reciprocal agreement pursuant to section 18 of the Act." The operational question is how that is proven from a compliance desk.

§ 1523.3(c) provides the answer and is where the real work is concentrated. The provider may rely on a representation from the foreign issuer itself, subject to two cumulative conditions: (1) it cannot rely on that representation unless it performs "reasonable due diligence" on it, and (2) it cannot rely on it if, based on that diligence or other reasonably available information, it "knows, has reason to know, or should know" that the representation is false. The standard is not good faith: it is imputed knowledge.

§ 1523.3(d) lists five non-exhaustive examples of what counts as an offer or sale to a person located in the U.S., and § 1523.3(e) describes the safe harbor that avoids that classification: reasonable belief that the recipient is outside the U.S., policies and controls designed to ensure this, and no advertising directed at the U.S. market.

The Treasury itself admits the flaw in this construction in the preamble: a provider "can never know with certainty whether a foreign payment stablecoin issuer 'will comply'" in the future with the terms of a legal order or a reciprocal agreement, and a strict reading of the provision would effectively shut down the offering of any foreign stablecoin. Hence question 54 of the file: whether due diligence should specify to what extent the provider must obtain and review the terms of the reciprocal agreement with the issuer's jurisdiction. The August 2026 memorandum from the U.S. law firm Sullivan & Cromwell describes the same timeline split: January 2027 for the foreign issuer prohibition, July 2028 for the general one.

Does USDT pass the legal orders and reciprocal agreement tests?

The two halves of the § 1523.3(b) requirement present very different levels of difficulty for the issuer of USDT.

The technical capacity to comply with legal orders is documented by the issuer itself. In its April 23, 2026 note, Tether announced the freezing of over $344 million across two addresses in coordination with OFAC (the U.S. Treasury's sanctions office), and cited cumulative statistics: cooperation with over 340 law enforcement agencies across 65 countries, more than 2,300 cases—over 1,200 with U.S. authorities—and more than $4.4 billion frozen in total, of which over $2.1 billion was linked to U.S. authorities. The freeze function exists, is used, and leaves a public on-chain trail: this half of the requirement is evidenced by signed documentation and verifiable transactions on the block explorer.

The second half is the one that does not depend on Tether. Reciprocal agreements under section 18 are created and implemented by the Secretary of the Treasury with jurisdictions that have comparable payment stablecoin regimes, according to section 18(d)(1) of Pub. L. 119-27; the variant that attributed this competence to the Federal Reserve for "substantially similar" regimes remained in the S.394 bill and did not make it into the enacted text. The issuer of USDT moved its headquarters to El Salvador in January 2025, after obtaining its provider license and stablecoin issuer permit from the National Digital Assets Commission; it was previously based in the British Virgin Islands. As of August 30, 2026, there is no record of any comparability determination by the Secretary of the Treasury under section 18(b) nor any reciprocal agreement signed with any jurisdiction, and the NPRM—the Notice of Proposed Rulemaking—mentions none.

It is useful to separate the 2027 path from the 2028 path, as they are different gates. For § 1523.3(b)—January 2027—technical capacity and willingness to comply are sufficient. For § 1523.3(a)(2)—July 2028—the foreign issuer must meet the four criteria of section 18(a): a home regime declared comparable by the Secretary of the Treasury under section 18(b), issuer registration with the OCC, reserves in a U.S. financial institution sufficient to meet the liquidity of U.S. customers—unless a reciprocal agreement allows otherwise—and domicile in a country not under comprehensive economic sanctions nor designated as a jurisdiction of money laundering concern. As of August 30, 2026, the first two do not exist for any jurisdiction in the world. In Tether's first full audit, published on August 19, 2026, the dress rehearsal for this examination was already visible.

Does Part 1523 prohibit holding USDT in a self-custody wallet?

The text proposed by the Treasury does not contain the expression self-custody a single time. It contains the concept, phrased differently, in § 1523.4(c)(3): "Any transaction by means of a software or hardware wallet that facilitates an individual's own custody of digital assets." The prohibitions in § 1523.2 and § 1523.3 do not apply to those operations.

The architecture of the rule is consistent with this. The regulated entity of the two provisions is the digital asset service provider—defined by section 2(7) of the law as anyone who, for compensation or profit, exchanges digital assets in the U.S., transfers them, safeguards them, or provides financial services linked to their issuance—along with the issuer. The individual holder moving funds from their account to their hardware wallet does not appear as the target of any prohibition.

There is a nuance that prevents reading this as a general shield. The preamble warns that the secondary market for unregistered stablecoins "at any time may implicate the prohibitions on offer and sale in section 3(b)," but it implicates the one who offers or sells: § 1523.2(d)(3) limits participation in an illicit issuance to assistance provided "at or around the point of issuance" and excludes secondary trading without a "close temporal nexus to the initial issuance." The holder feels this indirectly, when their exchange makes a decision.

Appendix A adds a piece of geography. Its first interpretation describes a U.S. resident on vacation abroad to whom a foreign issuer issues a stablecoin: there is no violation of § 1523.2(a), because the definition of "located in the United States" requires physical presence at the time of the operation. Location is measured by where the buyer is that day, not by their passport.

What is at stake in USDT, USAT, and USDC?

The three most relevant stablecoins fall into different boxes of the proposed Part 1523, and the size of each box is very unequal.

Own calculation. The two USDT figures often presented as contradictory are the same figure with and without treasury holdings. The issuer's transparency page publishes total tokens and the balance held in treasury by chain. Summing both sets of fields on August 30, 2026: 193,628,419,523 tokens issued minus 10,180,951,965 in treasury gives 183,447,467,558 in implicit circulating supply, compared to the 183,477,863,203 published that day by the independent aggregator. The difference is $30 million out of $183 billion: 0.017%. The starting assumption—that the treasury balance consists of minted tokens not yet put into circulation—is verified within that margin.

Own calculation. The implicit circulating supply of USDT deduced from the issuer's transparency page, divided by the USAT circulating supply published by the aggregator on the same day: 183,447,467,558 (issuer, Aug-30-2026) ÷ 184,151,715 (aggregator, same day) = 996.2. For every dollar of USAT, there are $996 of USDT. Comparing only issuer-side figures—Tether's implicit supply versus the 175,245,527 tokens from the USAT attestation as of July 31—the ratio is 1,046.8, or 1,047:1: the first version measures the same day with two different sources, the second measures two issuer figures separated by seven weeks. In either reading, the product Tether built to operate within the U.S. framework represents about 0.10% of the product targeted by § 1523.3(b). And USDC, with $74.007 billion, remains the real point of comparison for USDT in the regulated market—we analyze this in detail in the comparison between the two.

Why do the $10.181 billion in Tether's treasury count as future issuances?

The definition of "issue" in the proposed § 1523.1(c) is the piece that turns a balance sheet figure into a regulatory figure. Issuing is the first transfer of the stablecoin by the issuer that gives a third party the right to use, transfer, or redeem it. The preamble makes the consequence explicit: "a digital asset that has been minted but is held in the issuer's treasury would not be considered to have been issued as a payment stablecoin because the digital asset has not yet been transferred to a third party."

The following sentence closes the circle on the other side: after converting, redeeming, or reacquiring a stablecoin, "the first subsequent transfer of the payment stablecoin by the issuer ... is considered a new issuance, whether or not the transfer is characterized as a reissuance."

Note 30 of the NPRM confirms the reading from the outside: the Treasury observes that the OCC already proposed, in 91 FR 10202, 10208 (March 2, 2026), to define outstanding issuance value by excluding stablecoins held in the issuer's treasury. Both regulators treat the treasury balance the same way.

Own calculation. Applied to the August 30, 2026 snapshot: the $10,180,951,965 that Tether maintains in treasury is not inert stock under this rule; it represents future issuances waiting to happen. Every time one of those tokens leaves for a third party, it will be a new issuance, subject to the regime in effect that day. That reserve is equivalent to 55.3 times all existing USAT and 13.8% of all USDC in circulation: it is the ammunition that decides how much U.S. demand can continue to be served via the foreign route.

What is the pace of USAT according to its own attestations?

USAT is the product Tether launched on January 27, 2026, to operate within the U.S. framework: it is issued by Anchorage Digital Bank, National Association, a federally chartered national trust bank supervised by the OCC, which publishes monthly reserve attestations under the 2025 AICPA (American Institute of Certified Public Accountants) criteria for stablecoin reports. The series of seven signed reports explains better than any statement where the domestic plan stands.

The series offers three takeaways. First: the start was slow and the jump arrived in a single month, April 2026, when the circulating supply multiplied by 6.39, going from 22 to 141 million tokens. Second: after that jump, the pace fell to double-digit monthly percentages, +11.1% in May and +13.0% in June. The third is the most recent: between the June 30 and July 31, 2026 cutoffs, the USAT circulating supply dropped from 176,794,845 to 175,245,527 tokens, a 0.9% decrease, the first negative month since launch. The data was published by the issuer itself: the auditor signed the July report on August 28, and it appeared on the attestations page on August 31, one day after the data cutoff for this analysis.

An isolated month says little about a trend; what it does is withdraw empirical support for the idea of an ongoing domestic substitution, just when it should be accelerating. To absorb the U.S. demand currently served by USDT, USAT would need to grow by three orders of magnitude before January 2027.

What alternatives did the Treasury discard in the GENIUS Act NPRM?

The impact analysis under Executive Order 12866 lists the options the Treasury considered and rejected, and one would have changed the entire map: a longer transition—citing 36 months as an example—with a safe harbor to continue issuing unregistered foreign stablecoins in the U.S. at a minimum volume, with an example threshold of less than $1 billion in U.S. hands. It was discarded because the delayed benefits, particularly consumer protection, would outweigh the savings in transition costs. That threshold would have excluded USDT by two orders of magnitude in any estimate of its U.S. portion; what the rejection reveals is that the Treasury considered the gradual path and closed it.

It also rejected the opposite option—removing the safe harbors in § 1523.2(c) and § 1523.3(e) or the reasonable belief standard—due to compliance burden and the risk of migration to other jurisdictions. In the cost column, the document accepts two effects by name: "market concentration costs and switching costs" among the quantified ones and "reduced product choice" among the qualitative ones. Market concentration and fewer options for the user appear there as the assumed price of the rule. The file formulates 87 questions for public comment (own count of the full text), and twenty—from 51 to 70—fall on § 1523.3, the provision that starts in January 2027.

What can be verified now and what depends on the October 19, 2026 comment closing?

There are three things verifiable today against dated documents, without waiting for the final rule.

1. Which provision each held stablecoin depends on. If the issuer is domiciled outside the U.S., the provision is § 1523.3(b) and the date is January 18, 2027; if it follows the domestic path, it is § 1523.3(a) and July 18, 2028.
2. Whether the foreign issuer has already published any statement regarding its capacity to comply with legal orders. The continuation of the listing depends on that statement and the due diligence the exchange performs on it.
3. Whether the provider where the funds are held has said anything public about its plan. As of August 30, 2026, the comment period remains open and compliance desks are in the phase of reading the file.

The thesis of this article—that the rule acts on the provider, not the holder, and bites a year and a half earlier than the 2028 date suggests—would cease to hold in three scenarios: if the final rule removes or restricts § 1523.4(c)(3) and self-custody is no longer exempt; if it aligns the two dates, also moving § 1523.3(b) to July 2028; or if it extends § 1523.2 or § 1523.3 to the individual holder. All three would be visible in the final text, and none are in the proposed one.

This rule also does not overlap with European ones: the MiCA stablecoin register operates with a different logic and different timelines. Anyone holding USDT in both a European and a U.S. exchange is looking at two different calendars, and the one that arrives first is January 18, 2027.

Sources and links: Federal Register, 91 FR 53368 (Aug-18-2026) · Full text of the NPRM · Official PDF on GovInfo · Department of the Treasury press release (Aug-17-2026) · Tether, transparency.json · Tether, coordinated freeze with OFAC (Apr-23-2026) · Anchorage Digital, USA₮ reserve attestations · Circle, USDC reserve transparency · DefiLlama, stablecoins API · Sullivan & Cromwell, NPRM analysis