NEAR Protocol co-founder Illia Polosukhin stated Monday that AI-assisted hacking began last October and is now outpacing traditional code security reviews. Speaking in an interview on The Block's The Starting Block podcast, Polosukhin identified the Balancer incident as the first example of this trend. He attributed the acceleration to AI models' capability to find software vulnerabilities faster than human reviewers can patch them. The cybersecurity industry has historically relied on manual code reviews, but Polosukhin argued this approach has become unsustainable as AI tools evolve.
Polosukhin Proposes Formal Verification as Security Solution
Polosukhin advocated for formal verification that mathematically proves code "does what it says it does." He said this process has been historically expensive, slow, and labor-intensive, but AI can now help generate most of the necessary proof, which can then be independently checked.
"You don't need to rely on AI because the AI just produces the proof," Polosukhin said. "You just verify that the proof is correct."
He argued that cybersecurity up to this point has mostly involved "smart people" reviewing code and hoping that no one smarter finds something they missed. "We need a different way of thinking," Polosukhin said. "It's not like, 'Oh, we're going to review everything with a better model and hope that the next better model is not going to find it.'"
Polosukhin stated this approach should eventually go beyond smart contracts to critical internet and networking software. "We need to fundamentally shift to building blockchain code --- and honestly, all the code on the internet --- to be formally verified," he said.
Zcash Project Tachyon Demonstrates AI-Assisted Proof Generation
Project Tachyon used AI-assisted verification to produce a mathematical proof that Zcash's Ironwood shielded pool cannot contain the same kind of undetectable counterfeiting bug that was found in its Orchard pool last month. Project Tachyon said that advances in AI-assisted proof generation have reduced work that once took years down to a matter of weeks.
NEAR Rolls Out Quantum-Resistant Upgrade
NEAR rolled out an upgrade on Monday aimed at preparing the network for emerging security and scalability demands. Users can now switch to quantum-resistant account keys, while the network can automatically expand its capacity as activity grows.
FAQ
What did NEAR Protocol co-founder say about AI-assisted hacking?
Illia Polosukhin stated Monday that AI-assisted hacking began last October with the Balancer incident and is now outpacing traditional code security reviews. He said AI models can find software vulnerabilities faster than human reviewers can patch them.
How does formal verification work according to Polosukhin?
Polosukhin described formal verification as a process that mathematically proves code "does what it says it does." He said AI can now help generate most of the necessary proof, which can then be independently checked without relying on the AI itself.
What upgrade did NEAR announce?
NEAR rolled out an upgrade on Monday that allows users to switch to quantum-resistant account keys and enables the network to automatically expand its capacity as activity grows.
