The INFORM Consumers Act took effect in June 2023. Its purpose was to make online marketplaces verify high-volume sellers and make it harder to convert stolen merchandise into cash.
Three years on there has been exactly one federal enforcement action: a $2 million civil penalty against Temu in September 2025, which the FTC itself described as "the first action to enforce the INFORM Act."
Read what Temu was actually penalized for. Failing to provide a telephone number for reporting suspicious activity. Failing to offer any reporting mechanism inside its gamified shopping features until November 2024. Failing to make seller disclosures clear and conspicuous on mobile.
Not one element of the first-ever INFORM enforcement action concerned a stolen good.
A correction first: edition #17 reported that no INFORM enforcement actions had been brought, citing a letter from the law's own authors. That was wrong — the Temu action predated the edition by eleven months. The error is instructive, because the case is so far from what the law was sold as that it did not register as enforcement at all.
What three years produced
In January, Senators Durbin and Cassidy released data collected from 46 companies on what INFORM has actually done:
Roughly 40,000 high-volume seller accounts suspended for INFORM violations.
Roughly 20,000 of them later reinstated after coming into compliance — a 50% reinstatement rate.
Roughly 3 million non-high-volume accounts suspended or removed for "stolen, counterfeit, or unsafe goods" — three categories fused into one number that cannot be separated.
Roughly 24 million product listings removed.
Those are real numbers and they are not nothing. But note what is missing: no platform-specific breakouts were released. You cannot tell whether one company did nearly all of it or almost none. And the only category that names stolen goods lumps them with counterfeits and safety recalls.
The disclosure asymmetry
Compare how the two platforms that do publish enforcement data describe two different problems.
Amazon's 2026 Trustworthy Shopping Experience Report, covering 2025, on counterfeits: more than 15 million counterfeit products identified, seized and disposed of. 99.9% of suspected infringing listings blocked proactively. More than 32,000 bad actors pursued across 14 countries since 2020, producing more than 290 prison sentences averaging 29 months.
The same report, on stolen goods, in full: "Millions of suspected stolen products identified and seized."
That is the entire disclosure. No unit count, no method, no baseline.
eBay's 2025 Global Transparency Report blocked 244.5 million prohibited or restricted item violations, removed 7.9 million potentially counterfeit or prohibited items, suspended more than 55,000 accounts, and received 19,611 law enforcement records requests. It does not break out stolen goods anywhere.
Of the seven major resale platforms, not one publishes a stolen-goods-specific enforcement figure. Two publish counterfeit figures to the unit. The precision exists. It is pointed elsewhere.
A perspective from the field
We spent years arguing that the resale channel was the chokepoint, and we got a federal law. What we did not do is specify what the law should require anyone to count.
A platform can tell you it seized fifteen million counterfeits and, in the same document, that it seized 'millions' of stolen goods. One of those is a number and one is a word. Nobody made them count ours, so they didn't. That is not a platform failure. It is a drafting failure, and the industry that lobbied for the draft owns it.
The federal floor became a ceiling
This is the part almost nobody in the industry has tracked, and it matters more than the enforcement count.
California's SB 1144 took effect July 2025 and went well past the federal law. It required marketplaces to alert law enforcement when aware of stolen-goods sales to California residents, maintain written anti-ORC policies, provide public reporting mechanisms for suspected stolen goods, and open confidential channels to police. It reached off-platform transactions by high-volume sellers, and it let district attorneys and city attorneys enforce it.
It has never operated. NetChoice sued, and on July 11, 2025 a federal judge preliminarily enjoined the law as likely preempted — by INFORM's own express preemption clause — and barred by Section 230. Georgia's comparable statute was enjoined too.
Summary judgment briefing in the California case completed on September 3, 2026. It is live right now.
So the sequence is this: the industry won a federal law that imposed verification duties but no stolen-goods duties, and that law's preemption clause is now the instrument blocking the two states that tried to impose the duties the federal law left out.
What the prosecutions actually show
Two things, both of which cut against the standard narrative.
The architecture is physical intake with an online exit. In a federal sentencing in western New York in November, a pawn shop was found to have paid boosters $290,000 for merchandise carrying roughly $1.16 million in retail value — 25 cents on the retail dollar — across 37,936 items and more than 670 transactions, resold on eBay. An earlier New York case documented a fence paying 6 to 8%. That is a fourfold spread between two prosecutions, which means anyone quoting you a single fence margin is choosing one. But in both, the intake point was a physical business. The marketplace was the exit, not the entrance.
And the goods are ordinary. California's organized retail crime task force reports recovering 1,620,803 items worth $74.6 million since 2019. That is about $46 per item. Detergent, razors, over-the-counter medicine. It is a useful corrective to the transnational-syndicate framing, and it points your controls at the categories that actually move.
Three practical moves for the next 90 days
Stop treating the marketplace as the chokepoint. The documented architecture is a physical intake point — pawn shop, second-hand store, residence — feeding an online storefront. The intake point is local, enforceable, and the thing federal law does not touch.
Ask your platform partners for the stolen-goods number specifically. Not counterfeits, not "prohibited items," not a combined figure. If they cannot produce it, that answer belongs in your next executive briefing, because it defines what your referrals are actually worth.
Track NetChoice v. Bonta. If California loses, state-level stolen-goods duties on marketplaces are effectively foreclosed and the federal law is the entire ceiling. That changes where this industry should be spending its advocacy, and it will be decided while you are building next year's plan.
Closing note
The resale layer is the one part of this problem where an actual enforcement and compliance apparatus got built. Three years in, it has produced one federal case about a phone line, an aggregate number that fuses stolen goods with counterfeits, zero stolen-goods disclosure from any major platform, and a preemption clause blocking the states that tried to go further.
That is not an argument that the law was pointless. It is an argument that the industry specified the wrong thing, and has not yet noticed.
If you have obtained a stolen-goods-specific enforcement figure from any platform, I would like to see it. Reply with anything you can share, anonymized always.
Forward this to one LP or AP leader who should be reading it.
— Gabriel
The LP Brief is a weekly intelligence read for senior loss prevention and asset protection leaders. Free. No vendor noise.
Not yet subscribed? thelpbrief.com