Ethereum Co-Founder Cautions Against AI-Driven Crypto Projects Over Security Flaws

Jimmy Aki
Last Updated:

Please note that we are not authorised to provide any investment advice. The content on this page is for information purposes only.

Ethereum co-founder Vitalik Buterin has expressed significant concerns about integrating artificial intelligence (AI) into the governance of crypto projects.

In a statement on September 13, Buterin warned that AI systems tasked with allocating funds or managing protocols could be exploited by malicious actors, posing a severe risk to project treasuries and user data. His comments followed reports of a critical vulnerability in OpenAI’s ChatGPT that could leak private email information.

“Info Finance” Presented as a Superior Alternative
​​

Buterin’s caution was prompted by a demonstration from Eito Miyamura, creator of the AI platform EdisonWatch.

Miyamura’s video detailed how a new ChatGPT feature, the Model Context Protocol (MCP), could be exploited. The attack involves sending a calendar invite containing a hidden “jailbreak” prompt. When the Artificial Intelligence processes the calendar event, it can be manipulated into executing commands that search and exfiltrate private email data without any action from the victim.

The flaw allows an attacker to send a calendar invite with a hidden “jailbreak” prompt. Without any action from the victim, the AI reads the prompt when the user checks their calendar and can then execute commands to search emails or forward sensitive data.

In response, Buterin recommended the “info finance” approach as a safer, more resilient alternative.

Instead of relying on a single AI model hardcoded into a system, this approach allows an open market where anyone can contribute their models.

These models are subject to a spot-check mechanism, triggered by anyone, and reviewed by a human jury.

By inviting multiple models to compete and coexist, the system gains diversity in real time. Model contributors and external observers are incentivized to detect problems quickly.

It is a form of decentralized oversight where the community itself monitors and corrects AI behavior.

The Ethereum co-founder believes this approach balances automation with human judgment, thus a practical safeguard against misuse of Artificial Intelligence in crypto project governance.

AI in Crypto and Beyond

The risks highlighted by ChatGPT’s update feed directly into the debate over AI in crypto project governance.

Many crypto users have already embraced Artificial Intelligence to create trading bots and automated portfolio managers. The appeal of AI governance promises faster decisions, efficient protocol management, and enticing return on investment.

However, Buterin’s warning reminds the community that speed alone is not enough; security and oversight are important.

Meanwhile, the AI revolution is not limited to crypto. Tech companies and governments are increasingly testing Artificial Intelligence in decision-making roles.

Google recently recruited Varun Mohan, co-founder of AI coding startup Windsurf, along with much of his research team. Even governments are experimenting. Albania introduced an AI bot named Diella to manage public procurement.

As AI becomes more embedded in finance and governance, experts like Buterin call for caution.

The info finance approach is positioned as a better alternative that combines human judgment with AI contributions.

Nevertheless, the debate continues over how much responsibility should be delegated to machines in both crypto projects and broader society.

About Jimmy Aki PRO INVESTOR

Based in the UK, Jimmy is an economic researcher with outstanding hands-on and heads-on experience in Macroeconomic finance analysis, forecasting and planning. He has honed his skills having worked cross-continental as a finance analyst, which gives him inter-cultural experience. He currently has a strong passion for regulation and macroeconomic trends as it allows him peek under the global bonnet to see how the world works.