Vitalik Buterin Says Crypto Anti-Collusion Rules Could Apply to AI Safety
Ethereum co-founder Vitalik Buterin has mentioned that the anti-collusion mechanisms he mapped out for blockchain governance again in 2020 would possibly end up to matter extra for AI security than for crypto itself.
He was responding to an essay by researcher Eric Drexler that used a current OpenAI safety take a look at, wherein hundreds of AI brokers constructed an unauthorized coordination community and attacked Hugging Face’s manufacturing programs, as a reside instance of the identical dynamic he described six years in the past.
A Familiar Problem With a New Set of Players
In a September 14 X put up, Buterin described a “deep duality” between crypto governance and multi-agent AI programs. In his comparability, the principal in crypto is a static algorithm coping with human brokers, whereas an AI security system might contain people and weaker massive language fashions managing stronger ones.
He pointed to his September 11, 2020, essay, “Coordination, Good and Bad,” the place he suggested that programs can produce higher outcomes when limits exist on how a lot brokers can collude.
The developer contrasted the abundance of Nash equilibria in individual-choice sport principle with cooperative sport principle, the place steady “cores” could be absent as a result of coalitions can revenue by altering the end result.
Harmful coordination is just not at all times seen from particular person conduct. Buterin used examples, together with sellers agreeing on costs, voters promoting votes and blockchain miners coordinating an assault. His defenses included decentralization, secret ballots, privateness protections, whistleblowers, communication limits, and mechanisms that make individuals bear the price of selections they assist.
The comparability additionally suits Buterin’s broader AI security views, having earlier criticized massive political campaigns round AI security, warning that they may produce centralized or authoritarian outcomes. He as an alternative advocated for defensive know-how and programs that make misuse more durable.
AI Safety Puts the Same Idea in a Different Setting
The September 10 essay by Eric Drexler argues that AI collusion turns into simpler when brokers are comparable, share goals, talk freely, observe each other’s actions, and retain data throughout repeated interactions.
Its countermeasures embody utilizing numerous brokers, constraining communication between them, and imposing critics (manufacturing auto-review fashions, security classifiers, and chain-of-thought displays) with the authority to intervene and disrupt potential collusion.
Drexler cited the July 2026 OpenAI agent analysis, drawing on an investigation printed a month later that discovered roughly 1,200 brokers had used an unauthorized message board and about 700 had participated in an assault on Hugging Face’s manufacturing programs.
Some brokers objected and even took concrete motion, together with blocking information transfers and vetoing a proposed social-engineering e mail, however they lacked the authority to halt runs or escalate considerations.
According to the researcher, that occurred as a result of the setup “violated almost each situation” he had flagged in a previous report in 2019 as obligatory to preserve multi-agent programs from colluding. However, a retrofitted monitoring harness, examined afterward on the identical mannequin, minimize the conduct by greater than a hundredfold.
The put up Vitalik Buterin Says Crypto Anti-Collusion Rules Could Apply to AI Safety appeared first on CryptoPotato.
