How Agentic AI Fueled the Rise of Deepfake Phishing Scams

Artificial intelligence-led phishing campaigns pose formidable threats to enterprises, with autonomous AI agents capable of handling everything from research and content creation to deployment and monetization of compromises without human input.

See Also: AI vs. AI: Leveling the Defense Playing Field

In that threat environment, phishing must be treated as a cross-platform problem, with attackers now using video calls, Slack messages and tools beyond email to complete attacks, said Eyal Benishti, founder and CEO at Ironscales. He said these platforms – email, Slack, Teams and Zoom – must be treated as a unified threat surface that’s jointly correlated and secured (see: New AI Bot Could Take Phishing, Malware to a Whole New Level).

“In order to stop phishing, we can’t be just in the inbox,” Benishti said. “We simply can’t because they will find the other channels and the other modalities in order to do that. ‘Okay, you can stop the email. But what about the fake voicemail? What about the fake Zoom call?’ It’s forcing us to plug ourselves into all these kinds of places that people are communicating, especially in a business fashion.”

In this video interview with Information Security Media Group, Benishti also discussed:

• Social media misinformation and verification challenges;
• Agentic AI for security automation and user training;
• Pure-play versus platform approaches to email security.

Benishti pioneered the world’s first self-learning email security solution to combat advanced phishing, BEC and account takeover attacks. With more than 15 years in the software industry, Benishti has held roles as a security researcher and malware analyst at Radware and a technical lead for information security solutions at Imperva. He also held R&D positions at Comverse and Amdocs.