Deepfakes could drive largest crypto hacks in 2026
Cyberattacks involving real-time deepfakes, phishing, supply chain compromises, and cross-chain infrastructure vulnerabilities are likely to be behind some of the biggest crypto hacks in 2026, according to CertiK senior blockchain investigator Natalie Newson.
As Cointelegraph reports, the crypto industry has already lost over $600 million to hacks since the beginning of 2026. Much of this is linked to two thefts attributed to North Korean hackers in April. These include the $293 million Kelp DAO exploit, which occurred on Saturday due to a single point-of-trust failure in the LayerZero cross-chain protocol infrastructure, as well as the $280 million Drift Protocol hack.
Another attack, also linked to the DPRK, involved the use of AI for social engineering. On April 15, crypto wallet Zerion revealed that North Korean hackers used AI in a long-term social engineering campaign to steal around $100,000 from the company’s hot wallets.
According to Newson, in some respects, the rapid advancement of AI will only increase the risks facing the crypto market. She therefore recommends storing crypto assets in cold wallets and always verifying the authenticity of URLs and smart contracts.
AI as a defensive tool
At the same time, AI can be used not only for attacks but also for defense. Newson noted that increasingly convincing deepfakes, autonomous attack agents, and agentic AI are emerging — capable of independently identifying vulnerabilities in smart contracts, writing exploit code, and executing attacks at machine speed.
However, as Newson emphasized, AI can also become one of the most powerful tools for defense. The growing use of AI has led to a surge in bug bounty submissions — both valid and invalid. It has also been reported that Anthropic’s Claude Mythos model, which is said to be capable of identifying vulnerabilities in major operating systems, is already being deployed defensively by a limited number of tech companies.
When deepfakes emerged
Video deepfakes began gaining traction in the late 2010s, when the first neural network-based algorithms capable of realistically swapping faces in videos appeared. Initially, the technology was seen as entertainment, with videos featuring altered appearances of actors or politicians spreading widely online. However, by the early 2020s, deepfakes had moved beyond experimentation: generation quality improved, barriers to entry decreased, and tools became widely accessible. This gradually transformed the technology from a novelty into a tool that could be used in fraudulent schemes.
Today, deepfakes have become a powerful weapon for cybercriminals. While video calls were once considered a reliable method of identity verification, they no longer guarantee security. A person’s face, voice, and behavior can now be convincingly replicated in real time. Attackers use this in social engineering schemes — impersonating acquaintances, colleagues, or executives to trick victims into installing malicious software or granting access to sensitive data.
It is worth noting that in 2025, North Korean hackers stole at least $2 billion worth of cryptocurrency.
Latest Scam Alert News
- Forex
- Crypto