Ethereum Foundation AI lead Davide Crapis and Ethereum co-founder Vitalik Buterin have shared a new method to protect user privacy when interacting with large language models (LLMs).
They pointed out that every message sent to an artificial intelligence (AI) system triggers an API request. These requests raise basic concerns for both sides: users want privacy, and providers need security and reliable payments.
In a February 10 blog post, they explained that today’s systems make it difficult to balance those needs.
Did you know?
Subscribe – We publish new crypto explainer videos every week!
What is Uniswap? (UNI Token Explained With Animation)
They said, “We need a system where a user can deposit funds once and make thousands of API calls anonymously, securely, and efficiently”.
They argued that service providers currently rely on two poor choices. One option requires users to provide personal details, such as an email address or credit card. The other option uses on-chain payments for each request, which are slow, expensive, and easy to trace.
Their proposed setup requires users to deposit funds into a smart contract before making any requests. After that, each request can be made without revealing who sent it.
Zero-knowledge proofs and rate-limit tools help providers verify payments and block spam without seeing users’ identities.
They explain that the system also checks that a user has not spent more than their original deposit and that any refunds are accounted for.
Vitalik Buterin recently explained how he views the difference between decentralized finance (DeFi) and stablecoin products that focus on yield. What did he say? Read the full story.
