As agentic AI moves from experimentation into operational security environments, organizations are increasingly focused on where it can deliver real value, and where it may introduce risk. Adoption is accelerating, with 73% of cybersecurity teams already using or developing these capabilities, reflecting both growing confidence and heightened concern about how far autonomy should extend. Industry leaders, including AppOmni’s Director of AI, Melissa Ruzzi, offered their perspectives in Cyber Security Tribe on how to balance innovation with control. 

Agentic AI is proving most effective in areas that demand speed, scale, and consistency (such as triage, investigation support, and operational efficiency) while raising caution when applied to decisions with broader business impact or irreversible consequences. Ruzzi’s insights emphasize that the real risk lies not just in what AI decides, but in how it is implemented, reinforcing the need for domain expertise, controlled use cases, and careful testing. Ultimately, defining clear boundaries, governance, and accountability is essential to ensuring autonomous systems enhance security operations without introducing systemic risk.