TLDR: Vitalik Buterin tested personalized diet and exercise recommendations through local AI orchestration, drawing on more capable remote models. The experiment combined carefully written pr
The Ethereum cofounder Vitalik Buterin has tested a system to obtain AI recommendations while limiting personal data exposure. His experiment used health and travel information to generate personalized diet and exercise suggestions.
A local model coordinated requests to capable remote systems, drawing on their reasoning and knowledge. The AI privacy setup combined carefully written prompts, private payments through zkAPI, and Tor routing. These addressed different sources of identity leakage.
Vitalik Buterin said the recommendations benefited from remote input, although privacy protections still needed improvement. He also reported slow responses and a tradeoff between sharing less information and receiving useful advice.
Vitalik Buterin identified his local coordinator as Qwen 3.8 Flash Next, which called frontier models when needed. A skill file guided those calls, instructing the local system to disclose as little personal information as possible.
Doing a bit of a self-experiment.
Goal: use my personal health and travel data to provide personalized diet and exercise recommendations for me, using frontier models but in a way that avoids leaking to them any private information.
Strategy: use a local model (Qwen 3.8 Flash… pic.twitter.com/mi38E5jlfN
— vitalik.eth (@VitalikButerin) October 4, 2026
The first privacy layer addressed both prompt contents and writing style. The local model composed questions, reducing the risk that remote services could identify him through phrasing or personal details.
The second layer covered payments, which can connect AI requests to an identifiable customer account. For this, the experiment used zkAPI, a system designed to separate payment authorization from user identity.
The Ethereum Foundation described zkAPI in an October 1 announcement as private usage credits for paid services. Users fund a vault, then authorize spending with zero knowledge proofs rather than revealing which deposit paid for requests.
In runtime key mode, temporary API keys cap spending, while signed usage receipts determine the actual charge. The payment service checks funding proofs without receiving the prompts sent directly to the AI provider.
Tor supplied the third layer, targeting network information such as IP addresses. Vitalik Buterin accessed zkAPI through a Tor-wrapped command line tool, combining payment privacy with network routing.
The design treats these protections as complementary because each covers a different route to identification. Removing names from prompts still leaves payment records or network details as potential links.
The Foundation also cautioned that AI providers can still read submitted prompts. Its documentation says repeated personal details, reused conversation histories, and writing patterns can allow separate sessions to be linked.
Vitalik Buterin reported that the experiment returned recommendations improved by knowledge from frontier models. However, he identified four weaknesses, spanning network design, request construction, local performance, and the balance between privacy and usefulness.
He argued that Tor handles individual request separation poorly and may provide insufficient privacy for this use. He also estimated latency was 10 to 100 times higher than it could be.
The skill file needed better strategies for deciding what information remote models should receive. That makes request preparation another unresolved part of the experiment.
Vitalik Buterin said Qwen generated roughly 20 to 30 tokens per second on his setup. He wanted speeds above 100 tokens per second before the system would feel comfortably fast.
The AI privacy tradeoff also remained visible: withholding more context reduced the help available from remote models. That tradeoff suggests payment and network safeguards alone cannot preserve personalized recommendation quality when requests omit important contextual details.
The experiment follows a September 27 essay in which Vitalik Buterin described Ethereum as a future cryptographic world computer. His vision combines blockchain security with cryptographic privacy, verification and decentralized computing outside the chain.
In that essay, he identified Hegotá, planned for next year, as the likely last normal fork. Later development would involve recursive STARKs, automated formal verification, optimized consensus and quantum safe technology.
He also cited PeerDAS as an early step toward this broader architecture. The intended result is cheaper, more scalable and more private computation secured through modern cryptography.
The post Vitalik Buterin Tests AI Privacy Through zkAPI and Tor Routing appeared first on Blockonomi.
