2026 Cybersecurity Predictions: AI-Driven Attacks, Adaptive Malware and Deepfake Scams

Summary

The speaker predicts that AI agents will automate cyberattacks, malware may adapt to evade detection, and deepfakes will fuel scams. They urge defenders to learn how to build and deploy AI agents.

Key points

• The speaker predicts AI agents could automate attack stages including scanning for weaknesses, gaining access, moving laterally, escalating privileges and exfiltrating data.
• They forecast that malware could use AI to diagnose problems, navigate target environments and change its code to evade investigation or containment.
• They warn that deepfake videos, voices and messages could be used for scams and credential theft.
• The speaker recommends that defenders learn AI-agent frameworks and how to deploy and scale agents to support organizational defense.
• They predict that managing fleets of AI agents will become an important part of cybersecurity work.