Vitalik Buterin called "AI governance" a "bad idea".

Co-founder of Ethereum, Vitalik Buterin, believes that using artificial intelligence (AI) for governance is a "bad idea." He warns that if AI is used to allocate funds, the system will be easy to exploit by jailbreak codes aiming to seize all the money.

Buterin's comments came after Eito Miyamura, CEO of EdisonWatch, pointed out a serious vulnerability in ChatGPT after the platform fully integrated the MCP protocol, allowing AI to access data from applications like Gmail, Calendar, and Notion. Miyamura stated that an attacker could steal all personal information with just one email containing a malicious appointment invite.

Instead of giving full power to AI, Buterin proposed an "info finance" model – an open market where multiple developers provide different AI models. These models can be randomly tested and evaluated by a panel of people, with the support of LLM, to ensure diversity and transparency.

Buterin warns that the trend of "AI as a manager" poses serious risks, not only regarding traditional AI safety but also potentially causing "significant value losses" in the short term.