Ethereum co-founder Vitalik Buterin acknowledged that an AI analysis by Franklin Wang, head of Co-Invest, accurately identified him as the author of his anonymous revision of EIP-7503.
And ... we have a winner!
My method when writing the post in 2024 was: I wrote it in Chinese, used qwen2.5 locally to translate it to English, then manually fixed all the bugs in the translation.
Notice that the stylistic hints that his AI picked up on were intellectual habits…— vitalik.eth (@VitalikButerin) July 6, 2026
On June 22, Buterin launched a public experiment to test whether artificial intelligence truly threatens online anonymity. He invited participants to find a document related to Ethereum that he published several years ago.
There have recently been claims that AI text analysis will make online anonymity untenable.
So let me cannibalize a piece of my own anonymity to do an experiment.
At some point this decade, I wrote a published document of medium importance to Ethereum - I estimate ~200 to 2000…— vitalik.eth (@VitalikButerin) June 22, 2026
The winner was Wang. Using AI, his team determined that the programmer authored the anonymous revision of EIP-7503 from December 2024. During the analysis, they examined 27 texts and estimated the likelihood of Buterin's authorship at about 20%—a figure roughly ten times higher than that of the next candidate.
“The main signal was not his words, but his reasoning style,” Wang wrote after confirming the result.
Buterin explained that he deliberately tried to conceal his authorship: he wrote the document in Chinese, then translated it into English using Qwen 2.5 and manually corrected the text. However, AI was still able to identify him based on his explanations of mathematical and technical concepts.
“The stylistic features that the AI detected related to intellectual habits and the way algorithms were explained, rather than linguistic characteristics,” noted the Ethereum co-founder.
The experiment was prompted by a February report from researchers at ETH Zurich and Anthropic. The authors concluded that large language models are already capable of assisting in the de-anonymization of internet users through text analysis from open sources and related data searches.
Lighter CEO Vladimir Novakovski noted that in 2023, he and Wang used a similar approach to attempt to identify Bitcoin creator Satoshi Nakamoto based on the style of cryptographic publications. However, the researchers were unable to achieve a high-confidence result at that time. Wang later applied this method in Buterin's experiment.
As a reminder, at the beginning of July, the Ethereum co-founder presented the main priorities of the new roadmap for the second-largest cryptocurrency by market capitalization. According to him, developers will focus on quantum resistance, privacy, scalability, and restructuring key elements of the protocol.