Long a prime target for cyberattacks and systemic loopholes, the world of cryptocurrency is now facing an even greater threat with the rapid advance of artificial intelligence technology. Charles Guillemet, Chief Technology Officer at Ledger, has warned that recent developments in AI are shifting the balance in cybersecurity, intensifying risks for both crypto platforms and individual investors.
The role of artificial intelligence in cyberattacks expands
According to Guillemet, attackers are leveraging AI to discover system vulnerabilities faster and at much lower cost. Weaknesses in code that used to take months for seasoned security experts to uncover can now be identified within seconds through AI-powered tools. Guillemet emphasized the magnitude of this risk by noting, “It has become extremely easy to both identify and exploit flaws in a system.”
A series of cyberattacks in recent weeks bear out these concerns. Just this week, a major vulnerability in Drift, a decentralized finance protocol within the Solana ecosystem, was exploited for approximately $285 million worth of digital assets. Last week, the yield protocol Resolv suffered a loss estimated at $25 million. According to data from DefiLlama, the combined value of digital assets lost through cyberattacks or exposed system vulnerabilities has reached $1.4 billion over the past year.
Security flaws spread rapidly in the new AI landscape
Guillemet highlighted a critical point regarding security: ideally, breaching a system should be far more difficult and costly than the potential reward. However, with the integration of AI, this balance is being lost. The growing reliance of developers on AI-assisted coding tools is further accelerating the pace at which new vulnerabilities emerge and propagate throughout the crypto ecosystem.
He pointed out the scale of the problem, observing,
“There is no magic button that ensures safety. We’re going to see a massive output of code that is insecure by design.”
In Guillemet’s view, the crypto ecosystem must fundamentally rethink its approach to security protocols. Classic audit methodologies have their limits, and as the environment grows in complexity, stakeholders need to be acutely aware of these limitations.
He suggested that mathematical verification—formally proving code correctness—can be a more effective shield against vulnerabilities than traditional audit techniques. On the hardware side, strong solutions such as hardware wallets offer a critical layer of protection, with the ability to isolate private keys from internet access and thus minimize exposure to online threats.
Evolving malware trends pose additional challenges. Malicious software now routinely targets digital wallet keys on compromised mobile devices, sometimes siphoning assets without any user action whatsoever. This demonstrates how attack methods are adapting in sophistication as AI and automation become more accessible.
Guillemet cautioned that everyday crypto users should avoid assuming that systems in use are inherently secure. He advised,
“It’s wise to presume that most of the systems you use cannot be fully trusted.”
Looking ahead, security practices such as cold storage, stronger authentication procedures, and keeping sensitive data offline are likely to see wider adoption among crypto holders. Still, software solutions alone are insufficient—preparedness against both digital and physical threats is required. Guillemet believes that while core infrastructure, such as wallets and protocols, will gradually fortify their security, the broader software ecosystem could face significant challenges adapting to this new era.
