Levie argues that AI model proliferation benefits cyber defense

On October 25, 2026, Box CEO Aaron Levie put forward the view that as AI models proliferate, the application layer can quickly equip defenders with tools to protect their environments, which is beneficial for cyber defense. He said that in the future various types of agent clusters will emerge to scan, detect, protect against and handle the risks enterprises face, and that every layer of the security stack—identity, firewalls, network, code scanning, data protection and so on—represents a huge opportunity; the only thing lacking is investment in compute. In a report later the same day, Levie predicted that before long the number of tokens flowing through cybersecurity agents will be 1,000 times greater than now; as the attack surface expands, enterprises will struggle to keep up with the number of new risks, while agent clusters can invoke various security tools, test attack variants, read logs and assess enterprise data risks. He said this capability stems from the maturation over the past few months of long-running agents and computer use, as well as deep training of models in the cybersecurity field. At present this view remains Levie's personal judgment, with no further implementation progress yet.

AI-generated from public reports

1. Report date Oct 25· 2 reports

   Levie argues that the proliferation of AI models benefits cyber defense

   Source:Aaron Levie