The mathematics protecting Bitcoin and Ethereum could face an unexpected threat from artificial intelligence. Researchers are questioning whether AI might discover a shortcut that allows hackers to break the encryption and digital signatures protecting the internet and cryptocurrencies.
The concern follows OpenAI’s October 6 announcement that an unreleased AI model had produced 722 mathematical manuscripts, covering 372 groups of results.
The model attempted roughly 4,000 unsolved problems. The findings are still being evaluated.
The breakthrough prompted a warning from Johns Hopkins cryptographer Matthew Green.
Green later clarified that he knew of no imminent cryptographic breakthrough. His concern is that AI could find mathematical weaknesses that decades of human research have missed.
How Could AI Put Bitcoin and Ethereum Wallets at Risk?
Bitcoin and Ethereum rely on elliptic-curve cryptography to protect transactions. A complex mathematical problem makes it practically impossible to work backward from a public key to its private key.
If AI discovered a much faster way to solve that problem, attackers could potentially recover private keys and steal funds.
Bitcoin wallets with exposed public keys would be particularly vulnerable. Ethereum accounts that have signed transactions could also face risks.
The potential consequences extend to online banking and secure internet communications.
“What I would argue is that most of the open problems solved by OpenAI with their new models were actually rather close to a solution, conceptually speaking, and that mathematicians working in the field would have come to solutions soon with new tools in their hands.“
Modern cryptography is generally designed around extremely high security levels. Green suggested AI-driven discoveries could reveal that some systems believed to offer 128-bit security actually provide only 96 or 108 bits.
A reduction of 20 bits would make an attack roughly one million times cheaper. Even then, an attack might remain prohibitively expensive.
Increasing key sizes could help. But Green questioned how security experts would determine the right size.
“But crank them up to what?”
Even Post-Quantum Cryptography Could Face Questions
The concern also reaches the systems being developed to protect against quantum computers.
In 2024, the US National Institute of Standards and Technology approved new post-quantum standards, including ML-KEM and ML-DSA.
These rely on lattice-based mathematics. If AI finds substantially better attacks against those problems, some of the industry’s planned protections could require changes.
However, Blockstream CEO Adam Back called the warning “dubious.”
Blockstream researchers have developed SHRINCS, a hash-based signature system using SHA-256. Its compact signatures start at 324 bytes, and the technology has been demonstrated on Bitcoin’s Liquid sidechain.
Bitcoin would still require a protocol upgrade to adopt it. Hash-based signatures also cannot directly replace public-key encryption and key exchange.
Some investors are already taking precautions.
Charles Edwards, founder of Capriole Investments and a BeInCrypto Market Intelligence Expert Council member, said he has stopped using DeFi and reduced his number of wallets and accounts.
If AI compromises mainstream encryption in any way, there will be chaos.”
For now, there is no publicly demonstrated AI attack capable of breaking Bitcoin or Ethereum’s core cryptography.
The question is whether the industry can upgrade its security quickly enough if AI eventually discovers one.
