Why are some experts saying DeFi is becoming unsafe?

Manuel Aráoz warned that artificial intelligence tools are rapidly improving the ability to discover smart contract vulnerabilities. He argued that AI-assisted attack discovery is making decentralised finance increasingly dangerous.

How could AI make DeFi hacks more common?

AI systems can help attackers analyse large amounts of smart contract code much faster than humans. Security researchers say AI may accelerate vulnerability discovery, automate exploit research and improve phishing or social engineering attacks targeting crypto users.

Is there proof that AI is directly carrying out DeFi exploits?

Not yet, according to Meir Dolev. He said there is still limited public forensic evidence proving AI independently executes DeFi attacks, though AI is clearly helping attackers scale reconnaissance and exploit development.

Why is DeFi especially vulnerable to AI-driven attacks?

DeFi protocols are highly exposed because their smart contract code is public, transactions settle instantly and protocols are interconnected through composability. Attackers only need to find one exploitable weakness to drain funds.

What did Yu Xian mean by a “dual threat”?

Yu Xian warned DeFi now faces two simultaneous risks: AI-powered hackers and organised social engineering groups. He said attackers are combining technical exploits with psychological manipulation to target protocols and users.

What role does social engineering play in crypto hacks?

Social engineering tricks users or employees into revealing sensitive information, approving malicious transactions or compromising systems. AI tools can make phishing campaigns and impersonation attacks more convincing and scalable.

Why did DeFi security concerns intensify recently?

April reportedly became the worst month for crypto losses since February 2025 after a surge in DeFi exploits. Several security firms linked the increase partly to advances in AI-assisted attack methods and operational weaknesses.

Are smart contract audits still enough for DeFi security?

Many security experts now argue that periodic audits alone are no longer sufficient. They say protocols need continuous monitoring, real-time threat detection and stronger operational security practices to defend against rapidly evolving attacks.

Image for illustrative purposes only. Not a real photo.

AI security fears divide DeFi industry

Written by Jon CuthbertPublished May 28, 2026, 5:12 AM

Debate erupted across the crypto industry after Manuel Aráoz warned that “all of DeFi” should now be considered unsafe because artificial intelligence tools are rapidly improving the ability to discover smart contract vulnerabilities.

Aráoz’s comments followed a surge in decentralised finance exploits during April, which became the worst month for crypto losses since February 2025 according to several blockchain security researchers.

Yu Xian said DeFi protocols now face a “dual threat” from black-hat hackers using AI systems alongside organised groups specialising in social engineering attacks.

Yu argued that decentralised finance teams must urgently adopt advanced AI-driven security monitoring tools capable of detecting vulnerabilities across live codebases, operational systems and offchain infrastructure.

Meir Dolev said there remains limited public evidence proving AI directly executes DeFi exploits, though he acknowledged that AI significantly improves the speed and scale of attack discovery.

Dolev said decentralised finance remains uniquely vulnerable because smart contract code is public, funds move instantly and attackers only need to exploit one successful weakness to drain assets.

Despite the growing concerns, security researchers argued the sector is not fundamentally broken and said DeFi protocols can still improve resilience through continuous AI-assisted auditing, stronger key management, transaction simulation systems and real-time security monitoring.