Ethereum is expanding its privacy ambitions into artificial intelligence payments through a new system called zkAPI.
The technology aims to separate API payments from user identities when accessing paid AI services. It also seeks to reduce the amount of personal information exposed through conventional accounts and payment credentials. The Ethereum Foundation says the model could support more private access to AI tools and metered digital services.
The Ethereum Foundation developed zkAPI alongside the Open Anonymity Project. The system allows users to pay for AI models and metered APIs without revealing their identity.
Traditional API systems usually connect an API key with an account and payment method. That structure can also link individual requests with a persistent user profile.
The Ethereum Foundation said this approach can expose sensitive information through payment records and prompt histories.
Its zkAPI architecture instead separates payment authorization from identity. Users first deposit ETH or USDC into an Ethereum-based vault contract.
Those funds become a private balance that users can spend across supported services. The payment process does not repeatedly expose account details.
zkAPI uses zero-knowledge proofs for private payments
According to the Ethereum Foundation, zkAPI creates spending authorization directly on the user’s device. The process uses cryptographic notes rather than conventional payment credentials.
A zero-knowledge proof confirms that the user holds sufficient funds without exposing the private note itself. One proof can authorize either one transaction or an entire session.
The system also uses unique serial numbers called nullifiers to prevent double spending. Duplicate nullifiers reveal repeated spending attempts without identifying the user.
Ken Liu, a Stanford computer science PhD candidate working with Open Anonymity, described zkAPI as an extension of unlinkable inference.
He said the system also separates payments from the underlying AI request.
The architecture includes local interfaces for OpenAI and Ollama services. Existing applications, editors, and chat clients can access those tools through localhost.
That structure could reduce the amount of redevelopment required when adding private payment functionality.
Privacy could become increasingly important as AI applications rely more heavily on paid inference and metered services.
Prompts can contain personal information, business strategies, proprietary code or financial data. Linking those requests with persistent accounts can create detailed usage records.
zkAPI aims to break that connection by making payment credentials independent from user identity.
However, broader adoption will depend on developer support, infrastructure performance and demand for privacy-focused AI payments.
Ethereum advances scaling and quantum resistance plans
The Ethereum Foundation is also pursuing wider network improvements through its 2026 protocol priorities.
Those priorities cover network scaling, user experience improvements, and stronger Layer-1 security.
Ethereum developers have started introducing quantum-resistant cryptographic algorithms across the execution and consensus layers.
The long-term roadmap targets full quantum resistance by 2029.
Ethereum also aims to increase network capacity further during 2026. The gas limit target stands above 100 million.
Ethereum educator Anthony Sassano said in November that 180 million would represent an ideal annual target.
Ethereum raised its gas limit to 60 million during 2025 following the Pectra and Fusaka upgrades.
Another major network upgrade, Glamsterdam, is scheduled for October.
The upgrade combines Amsterdam on Ethereum’s execution layer with Gloas on its consensus layer.
Amsterdam will introduce parallel processing for transactions. That approach allows Ethereum to process multiple transactions simultaneously instead of executing them sequentially.
The change forms part of Ethereum’s wider effort to increase throughput while strengthening its underlying infrastructure.
Together, zkAPI and Ethereum’s protocol upgrades show how developers are targeting privacy, scaling, and long-term cryptographic security simultaneously.
