AI, Cross-Chain Vulnerabilities Emerge as Top Crypto Hacking Risks for 2026

AI-driven attacks and cross-chain vulnerabilities are emerging as the main hacking threats facing the cryptocurrency market in 2026, according to an industry assessment.

Cointelegraph reported on April 23 that Natalie Newson, senior intelligence analyst at security firm CertiK, expects real-time deepfakes, phishing, supply-chain attacks and cross-chain vulnerabilities to be the main drivers of major hacks going forward.

The cryptocurrency market has already suffered more than $600 million in losses this year. The damage has grown amid a string of incidents believed to be linked to North Korean hackers. Among the most prominent cases were a roughly $293 million hack exploiting an rsETH vulnerability at KelpDAO and about $280 million in losses at Drift Protocol.

AI-powered social-engineering attacks have also surfaced. Wallet service Zerion previously said hackers linked to North Korea used AI-based social-engineering tactics to steal about $100,000.

Newson said the pace of AI development is accelerating the sophistication of attacks in some areas. To prevent phishing, users should verify the authenticity of URLs and smart contracts, she added. She also advised using cold wallets, an offline storage method outside exchanges, to protect assets without exposing private keys.

At the same time, AI can also serve as a defensive tool. So-called agentic AI, which automatically detects smart-contract vulnerabilities, has recently emerged and is strengthening security response capabilities.

Regulatory efforts are also expanding. OCCIP, a cyber and critical infrastructure protection body under the U.S. Treasury Department, said it would expand a cyber-threat identification program targeting digital-asset companies.

As AI adoption grows on both the offensive and defensive sides, infrastructure security and users' response capabilities are likely to become key variables for the future of the crypto ecosystem.