Key Highlights
- Grayscale highlighted DATA, USD.AI and Pearl as blockchain projects targeting different parts of the AI infrastructure market.
- DATA plans to migrate to Sui with a $30 million investment from the Sui Foundation, according to earlier The Crypto Times coverage.
- USD.AI uses onchain pools to finance loans backed by GPU equipment and related cash flows.
Grayscale has identified three blockchain projects targeting AI training data, GPU financing and compute monetization.
In an October 23 research article published on The Stack, Grayscale Head of Research Zach Pandl highlighted DATA, USD.AI and Pearl as examples of how blockchain technology is being applied to AI infrastructure.
The projects apply blockchain to data licensing, GPU financing and the monetization of computing workloads.
DATA targets AI training data
DATA focuses on collecting, licensing, and tracking human-contributed data, including photos, videos, and voice recordings. The project aims to let applications license these datasets to AI developers and compensate contributors.
On October 12, DATA announced plans to migrate its protocol to Sui, with Walrus set to serve as its default storage layer. The move includes a planned $30 million investment from the Sui Foundation and a proposed migration of DATA tokens at a 1:1 ratio.
USD.AI connects crypto capital with GPU financing
USD.AI focuses on financing AI computing infrastructure through loans backed by GPU equipment and related cash flows. Its onchain pools allow capital providers to fund these loans and potentially earn yield.
The model is designed to give computing providers access to financing for expensive hardware, while capital providers take on the associated lending risks. Returns depend on loan performance and borrowers’ ability to meet their obligations.
Pearl combines AI computing with blockchain security
Pearl aims to combine AI computing workloads with proof-of-work blockchain security. Computing providers may earn PRL token rewards alongside revenue from AI-related computing services.
Grayscale reported that PEARL rose 580% over three months as of October 22, citing CoinGecko data. The price increase does not establish whether demand for the project’s computing services is sustainable.
In September, The Crypto Times reported that about 12% of Pearl’s stated 2.1 billion token supply was circulating, following a rally that took the token to a record price of $1.76.
Pearl’s verification model has also faced technical questions. A June preprint by researcher Abhinaba Basu argued that verifying matrix multiplication did not, by itself, establish that the computation originated from genuine AI workloads.
The concern is whether the protocol can verify that computations originate from genuine AI workloads, rather than merely checking their mathematical correctness.
Grayscale examines blockchain’s role in AI
Pandl’s analysis places the three projects within a broader overlap between blockchain infrastructure and AI services. Other projects, including Bittensor, NEAR and World, are pursuing different applications involving decentralized AI services, AI agents and proof of personhood.
In an August research report, Grayscale identified agent payments, verifiable records and decentralized AI as areas where blockchain networks could support emerging AI applications.
Grayscale has also expanded its focus on AI infrastructure through investment products. In September, the firm renamed its Bitcoin Miners ETF as the Grayscale AI Compute ETF (GCPU), reflecting a focus on companies involved in AI computing infrastructure.
Grayscale’s latest analysis highlights training-data collection, GPU financing and compute monetization as areas where blockchain projects are targeting AI infrastructure.
Whether these models gain traction will depend on adoption, practical utility, and their ability to compete with existing alternatives.
Also Read: JPMorgan, BofA, Citi & Other US Banks Restrict AI Agents; CZ & Lawmakers Flag Risks