Ethereum Foundation launches privacy payment system for AI APIs on mainnet

Traders Union
All News
Crypto
Ethereum privacy AI payments

Highlights

Ethereum Foundation launched zkAPI on mainnet to enable privacy-preserving, on-chain verifiable payments for AI APIs using ETH or USDC.
zkAPI uses zero-knowledge proofs and vault contracts to separate API usage from user identity while enforcing spending caps and preventing double spending through nullifiers.
The system supports existing AI tools via localhost APIs but lacks full network anonymity, and the protocol remains experimental per the project's GitHub.

How zkAPI works on Ethereum

As reported by The Block, citing the Ethereum Foundation, the Ethereum Foundation said in a blog post published Thursday that it builds zkAPI with the Open Anonymity Project as a way to separate API usage from a user's identity while keeping payments verifiable onchain.

Users first deposit tokens such as ETH or USDC into a vault contract on Ethereum, where the balance is recorded as a private note. When they want to spend those funds, software on the user's device generates a zero-knowledge proof showing the request is backed by a funded note, without revealing which note or which deposit provides the funds.

A zkAPI server then checks that proof and issues a temporary API key with a spending cap. The user sends prompts directly to the AI model provider with that key, and when the key expires, the server deducts the usage from the user's private balance.

The system also publishes a serial number, known as a nullifier, each time a payment is made for AI usage. If a user tries to spend the same balance twice, the duplicate nullifier reveals the attempt without exposing other information.

Privacy limits and broader use cases

The client is designed to work with existing AI tools by exposing standard OpenAI and Ollama APIs on the user's machine, allowing apps, editors and chat clients to connect through localhost. The Foundation lists AI chat and AI agents as the first use case, while also pointing to blockchain RPC queries, image and video generation, VPN bandwidth and machine-to-machine payments between AI agents.

The Foundation says zkAPI does not provide network anonymity on its own. Requests sent from a stable IP address can still be correlated by the gateway, and users seeking stronger privacy may route traffic through Tor.

Sessions can also be linked again through prompt content, including personal details, writing style or conversation history. The project's GitHub repository describes the protocol as experimental.

zkAPI implements the ZK API Usage Credits design that Ethereum Foundation dAI lead Davide Crapis and Ethereum co-founder Vitalik Buterin publish on the Ethereum Research forum on Feb. 11. The Foundation formed its dAI team in September 2025 to improve Ethereum's role as a settlement and coordination layer for AI.

In our earlier coverage of Zcash’s Orchard shielded pool flaw, we explained how a long-running bug could have enabled counterfeit ZEC to be created without a public trace, underscoring the unique diligence challenges in privacy-preserving systems. We also described the project’s emergency response—sealing the old pool with a turnstile and launching the Ironwood pool—and how markets repriced ZEC based on both the confidence shock and the remediation steps.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.