AI Not Triggering DeFi 'Hackpocalypse,' Dragonfly Partner States

Artificial intelligence has not triggered a widespread "hackpocalypse" in decentralized finance (DeFi), according to Haseeb Qureshi, a managing partner at Dragonfly. Qureshi presented data indicating that both the total value stolen and the median size of hacks have been declining when compared to figures from 2025. This trend suggests that initial fears of AI being leveraged to orchestrate massive DeFi exploits were largely unfounded.

Qureshi's analysis, shared this week, contrasts with earlier predictions that advanced AI capabilities would lead to an unprecedented wave of sophisticated hacks targeting DeFi protocols. The data points to a stabilization, and in some metrics, a reduction in the severity and frequency of breaches attributed to AI-driven attacks. This observation implies that the security measures and defensive strategies within the DeFi ecosystem have either kept pace with or outmaneuvered the potential threats posed by AI.

The argument is supported by a comparative look at hack statistics. While specific figures for the current year were not detailed, the assertion is that the overall landscape of DeFi security incidents has not escalated as dramatically as some had anticipated due to AI advancements. This suggests a more nuanced interaction between AI development and cybersecurity in the financial technology sector, rather than a simple escalation of threats.

Dragonfly, a venture capital firm focused on cryptocurrency and blockchain investments, has a vested interest in the security and growth of the DeFi space. Qureshi's statement, therefore, carries weight within the industry, offering a perspective that challenges the more alarmist narratives surrounding AI's impact on financial security. The focus now shifts to understanding the ongoing evolution of both offensive and defensive capabilities in the context of AI.