What The Antivirus And Anti-Spam Wars Reveal About Securing AI

AI security follows the same pattern as antivirus and anti-spam wars, according to Harsh Singhal, who leads AI and Data Security at Glean. Threats like indirect prompt injection and agentic exfiltration are not new. Defenses must shift from deterministic to probabilistic detection, as previous cycles showed. Teams resisting this transition risk being caught flat-footed.