Avalanche Founder Warns AI May Target Existing Crypto Software Flaws

Avalanche founder Emin Gün Sirer argued that concerns about AI breaking cryptography are overstated and that AI is more likely to find existing software and wallet bugs first. He said codebases such as Ripple's could contain vulnerabilities that AI detects more easily than humans. The debate began Oct. 7 when Ethereum Foundation researcher Justin Drake urged the industry to prepare for possible ECDSA risks.
On Oct. 7, Ethereum Foundation researcher Justin Drake urged defensive steps, such as moving funds to unused addresses so hashing hides public keys. He feared AI-assisted math could weaken ECDSA signatures in months. Coinbase's Yehuda Lindell rejected this; Ledger's CTO warned against hurried migration; Vitalik Buterin saw serious risk without urging immediate action.
Avalanche's Emin Gün Sirer replied Oct. 9 that AI would likely find existing code and wallet bugs before cracking ECDSA. He cited Ripple and other valuable networks with limited code review. He still backs backup plans, including hash-based options Avalanche is exploring, but says securing current software comes first. September crypto losses hit $766 million, and a security team showed AI spotting 85 critical flaws.
If Sirer’s assessment proves correct, AI-assisted bug hunting could push wallet providers, exchanges, and developers to audit code more aggressively, potentially reducing losses for crypto holders. If ECDSA concerns intensify, users may face pressure to migrate keys, and rushed changes could introduce new mistakes. The debate may also affect public trust in crypto infrastructure and influence how quickly firms adopt defensive tools.