Solana Foundation CISO warns AI is making crypto scams more convincing

Michael Coates, chief information security officer at the Solana Foundation, said AI is making crypto scams harder to detect, with the primary threat coming from social engineering rather than blockchain vulnerabilities. Coates previously served as CISO at Twitter and led security at Mozilla before joining the Solana Foundation this year. His role extends beyond protecting the foundation itself to include work across the Solana ecosystem and engagement with regulators on cybersecurity standards.

Many recent crypto breaches did not stem from smart contract bugs but from compromised credentials and fraudulent identities, Coates told CoinDesk. He expects this trend to worsen as AI capabilities improve, enabling attackers to generate realistic messages, fake voices, and convincing personas at scale. Fully spoofed phone calls using voices of people victims know could become routine.

Crypto's permanence creates particular vulnerability: if a user signs a bad transaction or shares a seed phrase, there is no financial institution to contact and no transaction reversal mechanism. Coates does not believe scams can be entirely prevented. He argued that even careful users will eventually fall victim to well-crafted cons, making layered security essential. If one security layer fails, another must still protect the user or funds.

Coates emphasized that industry success depends on making the secure choice the default option, rather than requiring users to be security experts. Organizations need protection against fake vendors, investors, or colleagues using AI-generated audio or messages to trick employees into approving transactions or granting access.