Fighting AI with AI: PANW Chief SIO Talks Cybersecurity & Adjusting to Cyberattacks
Why It Matters
AI‑weaponized attacks are now fast enough to outpace human response, forcing businesses to adopt AI‑driven security or face rapid data loss and operational disruption.
Key Takeaways
- •AI-driven attacks are 400% faster, demanding machine-speed defenses.
- •Palo Alto reduces 30B daily alerts to one human‑reviewed incident.
- •Secure browsing must block AI prompt‑injection threats in browsers.
- •Identity misuse fuels 90% of breaches; rapid credential anomaly detection essential.
- •Leaders must adopt safe AI tools to balance innovation and security.
Summary
At RSA 2026, Palo Alto Networks’ Chief Security Intelligence Officer Wendi Whitmore warned that AI‑powered cyber‑attacks are accelerating dramatically, with Unit 42 reporting a 400% year‑over‑year increase in attack speed—malicious actors can breach a network and exfiltrate data in under 40 seconds. She argued that traditional, manual defenses are obsolete and that organizations must “fight AI with AI,” deploying machine‑speed detection and response.
Whitmore highlighted several data points: Palo Alto’s Security Operations Center processes over 30 billion alerts daily, yet AI‑driven automation narrows that to a single incident requiring human review. Nearly half (48%) of all incidents involve browser activity, often through AI‑enhanced assistants that expose users to prompt‑injection attacks. Moreover, identity abuse underpins roughly 90% of breaches, as attackers leverage stolen credentials rather than brute‑force entry.
“Secure browsing capabilities that can spot and block AI‑driven prompt injection at scale are essential,” Whitmore said, illustrating how a simple recipe search via an AI companion can become an attack vector. She also emphasized the dual concept of “cybersecurity for AI” – safeguarding the models themselves – alongside “AI for cybersecurity,” which powers rapid threat triage.
The implications are clear: security leaders must prioritize AI‑enabled defenses that operate at machine speed while ensuring those tools are safe to use. Reducing friction, protecting identities, and securing web interactions will allow enterprises to innovate with generative AI without exposing themselves to unprecedented risk.
Comments
Want to join the conversation?
Loading comments...