Coinbase has found a way to regulate AI development using the technology behind cryptocurrency.
The company is seeking to patent a system to “track the origin of machine learning data via blockchain,” essentially recording all data coming and going to AI models throughout the lifecycle.
Coinbase technology notes data that contributed to the model, such as training information, user input prompts, and corresponding outputs. It uses a “middleware component” or a system located between the model and the user to automatically log all interactions on the blockchain.
Using blockchain for this purpose provides an immutable, transparent record of who contributed to the model, managing ownership establishment and AI use. Furthermore, the decentralized nature of blockchains prevents a single part from unauthorized claims for ownership over the model. This is especially useful for open source models that often involve many contributors.
“One or more individuals, such as creators, developers, data scientists, engineers, and other stakeholders, could contribute to the development of machine learning models,” Coinbase said in its submission. “However, contributions to machine learning models may not be captured, and in some cases the output of the model may not be tied to the contributor.”
Blockchain hype fell from the peak of cryptocurrency and NFT fuels four years ago, but the underlying technology still has many uses. For example, JPMorgan Chase seeks Tech's patents that use blockchain for data systematics and fact checking, Intel previously filed a patent application for a blockchain-powered security audit, and Sony's patents detailed how to use technology to detect deepfakes.
And as copyright infringement lawsuits accumulate against companies like Openai and humanity, addressing the strengths of blockchain could be extremely helpful in tracking who owns what.
