Two questions deserve a straight answer before anyone builds on this. Has a token like this been made before, and what stops someone larger from doing it tomorrow. Both are answered on this page, with the evidence rather than the assertion.
Post a bond asserting something is true, face a challenge window, lose the bond if a challenge succeeds. This runs live at scale today.
No protocol bonds capital against claims about registered marks, evidence of use, chain of title, or sublicense validity.
Proposed in 2017, this was almost exactly the general idea: a list whose entries are staked, challengeable, and voted on. Nearly all of them died.
Any competent investor or engineer will find the token curated registry history in ten minutes. Naming it first, explaining precisely why it failed, and showing which of those failure conditions apply here is a stronger position than hoping nobody checks. Three of the five do not apply. The other two are design constraints we build against, and they are addressed below.
Bonded stake secures a claim when the cost of corruption exceeds the profit from it. That arithmetic assumes misbehavior is detected and reported. In practice detection is the weak link: the penalty is designed carefully, and then the protocol has to invent a reason for anyone to go looking.
a16z's analysis of slashing economics lays out the penalty side in detail and then concedes the other half, noting that how to incentivize the actual detection and reporting of misbehavior remains largely unresolved.
In a token curated registry the only reason to challenge an entry is the bond you might win. That is a thin, purely internal incentive and it is why challenge markets went quiet.
Brand rights are the opposite. An acquirer running diligence, a licensee in a territory dispute, a licensor who suspects underpayment, and a litigator preparing an opposition all have their own money riding on finding a defective record. They were going to do that work regardless. The protocol does not have to manufacture the audit incentive, because the industry already pays for it.
Was this specimen in use on this date. Does this sublicense exceed its parent term. Does this figure match the connected source. Each has a correct answer a challenger can demonstrate. The failure that killed adChain, rejecting the New York Times over editorial history, needs an absent fact to converge on. Here there is one.
Free anchoring produces a dense, useful record whether anyone ever stakes. Bonding secures only the subset of claims where money turns on the answer. A token curated registry had nothing underneath the token, which is why it had to be worth something on day one.
Holders govern parameters: fee splits, bond bands, schema additions. They never govern the outcome of an individual dispute. That escalates to a domain qualified panel. This closes off the exact mechanism that produced the canonical registry failure.
| Capability | PROBANT | Story Protocol | Rightsline, Flowhaven | EAS | Token curated registries |
|---|---|---|---|---|---|
| Registered trademarks | Yes | No, creative works | Yes | Generic, no schema | No |
| Evidence of use, timestamped | Yes, free | No | Stored, not provable | Possible, unstructured | No |
| Sublicense chain validation | Yes, enforced | No | Tracked, not enforced | No | No |
| Royalties from real world sales | Yes, connector sourced | Onchain derivatives only | Yes, self reported | No | No |
| Counterparty can verify | Yes | Yes | No, single sided | Yes | Yes |
| Capital at risk behind a claim | Yes, bonded | No | No | No | Yes |
| Buyer needs to hold a token | No | Gas in IP | No | No | Yes |
| Free tier that works alone | Yes | No | Enterprise contract | Yes | No |
Creative works: music, art, writing, software, AI training data. Its royalty module splits revenue between an original and its derivatives, which is a remix economy. It does not handle registered trademarks, royalty reporting from real world sales, or sublicense chains in commercial licensing programs. Reporting also notes that onboarding large IP holders has been slow. Different asset, different buyer, different money.
Serious software with serious customers. Rightsline alone reports more than 135 million IP assets under management for Fox, Sony, NBCUniversal, Activision, Scholastic and others. They replace spreadsheets with a central repository and compute royalties with thousands of exceptions. Every one of them is a system of record for one licensor. The counterparty cannot verify anything inside it, and the licensee's sales figure arrives as a number the licensee typed.
The general purpose attestation and schema layer, deliberately without a token. We use the pattern and add what a general primitive cannot supply: domain schemas that make a claim falsifiable, and capital standing behind the ones that matter. Not a competitor and worth building on rather than around.
Every free anchor is a record a competitor has to re-earn from scratch, and the anchor with the earliest timestamp is the one that matters evidentially. First mover advantage in a registry is unusually literal.
Built by a brand agency working since 2013 and onchain since 2020. The party that already builds the marks and manages the asset libraries is the party that can put a registry in front of the people who own them.
Evidence and administration, never legal effect. Teams that cannot resist claiming their chain creates or replaces a right lose general counsel in the first meeting, and general counsel is the buyer.
The proposed exemptions carry a US nexus test that the last cycle's offshore structures fail. Passing it is not a strategy, it is an accident of being a real company in the country where the rules are being written.
If nobody challenges, bonds secure nothing and the whole security model is decorative. Early attestation will be concentrated among a handful of participants, which is a security assumption rather than a decentralization claim. Connector coverage starts narrow, so most revenue figures will be marked as self reported rather than laundered into looking verified. And the regulatory path is a proposal, not a rule. Any of these can break this.
The mechanism, yes. The application, no. Bonded assertion with a challenge window and slashing runs at scale today in optimistic oracles such as UMA, in staked arbitration such as Kleros, in decentralized oracle networks such as Chainlink, and in restaking systems such as EigenLayer.
Applying that mechanism to registered brand rights, with license terms, sublicense trees, and royalty settlement as the records being attested, has not been done. Probant leads with the precedents by name rather than hoping nobody checks.
Token curated registries, the closest general precedent from 2017, asked token holders to vote on whether an entry belonged in a list, and paid them in the token whose price depended on the list looking good. Probant asks a bonded professional to attest to a specific verifiable fact, and pays yield in USDC out of collected fees.
The failure mode of a TCR was that curation was unpaid speculation and nobody had a reason to challenge a bad entry. Probant's challengers are the parties who already pay for diligence and royalty audits.
Three reasons, and each one has a specific answer in Probant's design. Curation had no external demand, so the list was worth only what speculators believed. Detection was unfunded, so bad entries went unchallenged. And rewards were paid in the registry's own token, which made the incentive circular.
Probant's registry is used by parties who need it whether or not a token trades, detection is funded by existing diligence and audit budgets, and yield is paid in USDC out of fees. The failure conditions are named because a design that cannot state how it fails is not a design.
Density. Every free anchor is a record a competitor would have to re-earn from scratch, and a chain of title is only useful if the upstream links are already in the same system. Registries compound, which is why year four is harder to displace than year one, not easier.
The second layer is position. Probant is built by a brand agency that has sat between brand owners and the systems governing their use since 2013, which is a distribution channel a protocol team normally has to buy.