A debate over AI security and digital control has grown after Ethereum cofounder Vitalik Buterin warned that people could lose control of their online accounts and personal data as AI systems become more powerful. He called for “self-sovereign identity, data and money,” saying users should keep direct control of their digital lives instead of relying on third-party platforms or automated systems.
The warning comes as concerns circulate online about AI tools handling sensitive account functions. A researcher known as Impulsive alleged that Meta’s AI support system could change Instagram account settings without proper identity checks.
In a recent blog post titled “My self-sovereign / local / private / secure LLM setup,” Vitalik laid out a different approach to artificial intelligence. He said many AI systems focus on convenience while leaving privacy and security exposed. He also warned that cloud-based tools could make it normal for people to share sensitive personal data with large centralized providers.
To reduce those risks, Buterin described a setup built around local AI models, offline data storage, and strict system isolation. He pointed to risks such as data leaks, manipulated prompts, hidden software backdoors, and accidental exposure of private information. As a result, he said privacy and security should be treated as basic design requirements, not optional features.
Buterin also described the tools behind his setup, including local language models, private data stores, and sandboxing systems that limit what AI agents can access. He said he also built communication tools that let AI read messages but require human approval before sending anything to other people.
He extended the same idea to cryptocurrency wallets. In that model, AI could help users prepare transactions, but humans would still approve higher-risk actions. He summed up this approach by saying, “the new ‘two-factor confirmation’ is that the two factors are the human and the LLM.”
Investor Saito linked Buterin’s views to what he calls “sanctuary technologies.” In a recent podcast discussion, Buterin said the bigger risk from AI is not just technical failure, but people slowly handing over decisions to systems run by companies or centralized intelligence.
He said technology should expand user choice rather than replace it. In his view, Ethereum represents an alternative system that people can adopt or leave at will. He also encouraged developers to build tools that protect privacy and keep decision-making in the hands of users.
