Mentioned in This Article
1. Why is DATA leaving its own blockchain for Sui?
The DATA Sui move marks the end of a company-run blockchain. DATA, the a16z crypto-backed project formerly known as Story Protocol, announced on October 21, 2026 that it will shut down its own network and relocate to Sui, following a $30 million investment from the Sui Foundation. The company says the change lets it concentrate on tools that verify where AI training data originates.
The decision caps several years of attempts to put intellectual property onto a blockchain. The company raised $140 million and built a dedicated chain for the purpose, yet found that reshaping how intellectual property is licensed was a harder task than it had expected. Chief executive Andrea Muttoni summed up the early ambition with the remark that the team was trying to boil the ocean, and that it was a very saturated and entrenched ocean.
2. How did Story Protocol become DATA?
Founded in 2022 in Bellevue, Washington, the project began with a wide mandate. The idea was that owners of music royalties, patents, trademarks and other creative or industrial assets could register them on a blockchain and attach usage terms that software could enforce. Licensing would then happen through open, programmable rules instead of negotiation between lawyers.
Reality proved less welcoming. Muttoni said many of the target industries already had established systems for rights and royalties, and that large rights holders showed little appetite for permissionless licensing. He also acknowledged that the team had been spreading itself too thin across too many kinds of intellectual property.
In June the company renamed itself from Story to the DATA Foundation and turned its attention to AI training data. Four months later it has narrowed the mission again, to tracking the origin and usage rights of that data. Muttoni presents the Sui migration as a continuation of the June rebrand rather than a fresh change of direction.
3. What does a dedicated chain cost, and what is Trace?
After the rebrand, the team asked whether it still needed a full blockchain of its own. Most of its work now centers on Trace, a registry that records the origin of a data set, the people who contributed to it and the usage rights granted to buyers. A registry of that kind does not necessarily require its own base layer.
Operating a layer-1 network is expensive. It requires engineers to maintain the software, validators to secure it, security audits and infrastructure providers. Muttoni framed the question directly, asking whether running a layer-1 still made sense or whether the energy should go into the protocol, the product and Trace. The answer was to close the chain and build on an existing network.
Muttoni did not say how much of the $140 million raised is left. He said existing investors, a16z crypto among them, remain involved in the company.
4. What problem is DATA trying to solve in AI data?
The new direction grew partly out of Poseidon, an AI data company that the Story team incubated and which collects and processes training data. Through Poseidon, the team saw a recurring difficulty: buyers often had no reliable way to confirm whether a data set was original, properly licensed or truly exclusive. In one episode, vendors offered supposedly exclusive data sets that other sellers had already sold. Muttoni said Poseidon was at one point offered its own data back.
Trace is built to address that gap. When a user uploads a video or a voice recording through a participating marketplace, Trace issues what amounts to a receipt. The record captures the time of submission, the terms the contributor agreed to, and any later licensing or payment events attached to the file.
Much of the early activity comes from Kled, a marketplace that pays people for their contributions. Muttoni says Kled adds three to four million contributions a day, a figure that includes older records being added to Trace. Those numbers measure registrations rather than sales, and DATA has not disclosed how many records have generated licensing revenue.
5. Why does the origin of training data matter?
Machine-learning models are only as dependable as the data used to train them, and the legal standing of that data has become a central question for developers and buyers alike. A company that licenses a data set needs to know that the people who supplied it consented to its use, that the seller actually holds the rights it claims to be selling, and that the same material has not been sold to a competitor under an exclusivity promise.
Today those assurances usually rest on contracts and on the seller's word. Disputes can surface only after a model has been trained, when removing the questionable material is costly or impractical. A tamper-resistant record created at the moment of submission, listing who contributed, when, and on what terms, is meant to move the verification step to the beginning of the process.
Public blockchains are a natural fit for such records because the entries are timestamped and cannot be quietly rewritten by one party. That property is the reason a registry like Trace can be useful to marketplaces, contributors and buyers who do not trust one another. It is also why the choice of network matters less than the quality of the data feeding the registry: a receipt is only as valuable as the confidence buyers place in the marketplaces that issue it.
6. What does the pivot say about crypto's search for demand?
DATA's path illustrates a pattern familiar across the industry. Projects launched with ambitious visions of rebuilding whole sectors onchain, then discovered that incumbents had little reason to switch. Narrowing to a specific problem with a clear customer, in this case buyers of AI training data who cannot verify what they are purchasing, is a more modest goal but a more testable one.
The shift also reflects the economics of running independent networks. A chain that has few applications still has to pay for security and maintenance, and the cost is the same whether activity is high or low. Migrating to a shared network turns a fixed operating burden into a dependence on someone else's infrastructure, a trade many small teams now judge worthwhile.
For the Sui ecosystem, the arrangement adds a project with a defined use case in a field that is drawing attention. Because the figures disclosed so far describe registrations rather than revenue, the true test will be whether provenance records translate into paid licensing and whether buyers come to require them.
7. Is DATA building a marketplace or infrastructure?
For now, DATA wants to act as infrastructure rather than as a competitor to the marketplaces that use it. Muttoni described the main goal as being the infrastructure for the auditability and provenance of data. He did not rule out running a marketplace of its own someday, but offered no details about timing or design.
The company is not alone in pointing toward artificial intelligence. A growing number of crypto projects are moving in that direction, and Muttoni estimates that the market for AI training data could be worth hundreds of billions of dollars. That is his estimate rather than an audited figure, and the market's size will depend on how widely buyers adopt provenance checks.
8. What does Sui add to the plan?
Relocating to Sui removes the burden of operating a network and supplies tools that match DATA's next steps. Walrus, a decentralized storage platform in the Sui ecosystem, could hold copies of the original files registered through Trace. Seal, another Sui component, could encrypt those files and restrict access to buyers who are permitted to see them.
That would address a current limitation. Trace records information about files but not the files themselves. With stored originals, a buyer could compare the data it receives against the version first submitted and detect whether anything had been altered along the way.
The Sui Foundation will become the largest institutional holder of DATA Network through its $30 million investment, although the financial terms were not disclosed. Muttoni said Sui takes a hands-off approach to product plans, leaving the company free to set its own roadmap.
Pairing storage with encryption also changes who can see sensitive material. Voice recordings and videos submitted by individuals are personal by nature, so a design in which files are encrypted and unlocked only for approved buyers fits the consent terms that Trace records. Whether those components are adopted in the way described will depend on engineering work the company has yet to schedule publicly.
9. What happens to the existing token and what comes next?
DATA will migrate its existing token to Sui on a one-for-one basis. The current chain stays live until the migration is finished. According to a press release, holders keep their rights and their vesting schedules through the transition, which means that locked tokens continue to unlock on the same timetable.
Muttoni described the shift as a matter of following demand. The team had been heading in one direction, he said, but the wind kept pulling it in a different one, and the part of the business that attracted the most interest is the one it now plans to build around.
The next milestones to watch are the completion of the token migration, any rollout of Walrus-backed storage for Trace records, and whether registrations from partners such as Kled convert into licensing revenue. The company has not yet published dates for any of them, so progress will be easiest to judge from future announcements.