The rise of artificial intelligence is significantly altering the landscape of cybersecurity, according to Sanaz Yashar, CEO of Zafran Security. In a recent interview, she revealed that the average time-to-exploit (TTE) has reached a historic low, with many vulnerabilities being weaponized before they are even patched. For the first time, analysts have recorded a negative TTE, indicating that attackers are exploiting bugs a day prior to vendor fixes.
Yashar, who co-founded Zafran in 2022, highlighted that 78 percent of vulnerabilities are now being weaponized using AI technologies. Her experience spans over 15 years, including a role within Israel's elite cyber group, Unit 8200, and positions at Mandiant and Google. She noted that while AI is not creating sophisticated attacks independently, it enhances the capabilities of threat actors, enabling them to execute breaches more quickly.
Furthermore, Yashar expressed concern over the expanding attack surface as organizations increasingly incorporate AI into their systems. This trend not only allows for prompt injection and other exploit techniques but also raises the risk of vulnerabilities within AI frameworks themselves, potentially leading to significant collateral damage.