Experts Assessed Risk of AI-powered Crypto Wallet Hacks

Published: October 11, 2026 – 11:04 UTC

An Ethereum Foundation researcher suggested that AI could crack crypto wallets before quantum computers.
Incrypted’s editorial team gathered experts’ views on how likely this is.
They noted that while AI-driven cryptography-breaking has not yet been proven, blockchains need a plan in advance to migrate to safer cryptographic mechanisms.

Cybersecurity and blockchain industry experts assessed the risk of crypto wallets being hacked due to a potential breakthrough in mathematical research enabled by artificial intelligence. The discussion intensified after Ethereum Foundation researcher Justin Drake said that AI could undermine cryptographic protections before quantum computers do, and urged investors to prepare for “bunker mode.”

Vitalik Buterin agreed that such risks need to be taken into account, while also urging people not to panic. Incrypted’s editorial team asked experts how realistic this scenario is, and how the blockchain industry can prepare for it.

How Could AI Potentially Hack Crypto Wallets?

Drake warned about the possibility of new mathematical methods emerging that would allow attackers to recover private keys from public ones. This would put digital signatures at risk, which confirm a user’s right to manage crypto assets.

In particular, this concerns ECDSA, which is used by, among others, bitcoin and Ethereum:

“This is the foundation on which both the bitcoin and Ethereum protocols are built. The foundation that, when a transaction is made, confirms that you, as a user, have the funds you are going to spend in that transaction. The potential (though for now rather hypothetical) vulnerability is that an attacker could impersonate the user and spend the user’s money for them,” LTV Protocol co-founder Andrii Sobol explained in a comment to Incrypted.

Hacken’s Offensive Security Services Director Grzegorz Trawiński emphasized that this is not about ordinary password guessing or speeding up existing attacks. This scenario would require a major breakthrough in cryptanalysis that breaks the mathematical assumptions on which digital signatures are based.

Sobol drew attention to the results of OpenAI’s mathematical research. According to him, the company published a large body of results from the new model across various areas of mathematics, but there was little material directly related to cryptography.

“The hypothesis is that OpenAI achieved no less impressive results, but simply did not publish them,” Sobol explained, noting that the lack of public results could indicate either an absence of major breakthroughs or a reluctance to disclose potential vulnerabilities before addressing the issues.

He noted that Tezos co-founder Arthur Breitman also pointed to this, drawing a parallel with the Manhattan Project, when scientists suddenly stopped publishing research on nuclear fission.

A tell from the Manhattan Project was that scientists who had been publishing about nuclear fission suddenly stopped publishing. As Justin points out here, the lack of cryptographic results in OpenAI’s mathematical breakthrough could itself be a tell. Evidence against it would be if they published a few improvements that aren’t massive breakthroughs.
— Arthur B. (@ArthurB) October 7, 2026

Do Cryptoassets Need Protection Right Now?

To protect their assets, Drake urged people to move cryptoassets more often to new wallets whose public keys have not yet been revealed.

At the same time, Travinsky believes there is currently insufficient evidence to justify an immediate switch to new wallets or key rotation after every transaction. According to him, no practical attack on the ECDSA algorithm has been demonstrated so far:

“It is important to distinguish between the severity of a hypothetical event and the evidence for how likely it is to occur. A practical compromise would have extremely serious consequences, but that does not mean it is inevitable in the near term.”

Quantus Network founder Yuvi Lightman noted that the threats from AI and quantum computers are interconnected. In his view, AI can both strengthen and undermine cryptography. He emphasized the importance of cryptographic agility — the ability to quickly update algorithms as new threats emerge.

Lightman also separately highlighted quantum computing. According to him, if the US quantum hardware and software company IonQ is not overstating its technological progress, the quantum memory resources required to break ECDSA could become available as early as 2027.

In the event of such a breakthrough, Lightman believes blockchain projects will split into two groups: those without post-quantum protection and those that have it in some form. Capital, Lightman said, could move from the former to the latter, and some investors might leave the crypto market altogether.

Travinsky, for his part, urged people to prepare for potential threats, but not to treat them as an incident that has already happened. Among the priorities, he cited assessing cryptographic dependencies, determining the extent of public key exposure, and developing plans to migrate to other algorithms.

According to him, premature migration could also create additional risks — from permission configuration errors to losing access to funds. Therefore, decisions about a large-scale transition should be based on compelling evidence of a cryptanalytic breakthrough, not just alarming forecasts.

Andrii Sobol called the likelihood of a hack of the cryptographic foundations on which certain blockchains are built in the coming years “low, but not zero.”

“That is why all blockchain protocols should think about potential recovery mechanisms that could provide a way out of a catastrophic situation with minimal losses,” he concluded.

Сообщение Experts Assessed Risk of AI-powered Crypto Wallet Hacks появились сначала на INCRYPTED.

Source: Incrypted