When queried about where their firms intend to allocate the majority of their resources over the coming two years, 43% of those surveyed selected AI agents. This figure is more than triple the proportion selecting API integrations, which was the second most frequently cited area.
A significant divergence among security professionals was observed regarding their capability to establish and sustain baselines for typical AI agent conduct. Close to half of the respondents indicated uncertainty in this area.
This finding suggests a critical operational challenge for infrastructure and platform security teams.
The emphasis on monitoring AI agents reflects a broader industry pivot toward managing risks associated with increasingly autonomous software components.
Why it matters
This trend signals a critical shift in infrastructure security focus toward monitoring autonomous software agents. Operators must prepare for the challenge of detecting malicious activity when the definition of 'normal' behavior is still evolving. Security investments are clearly prioritizing agent-level visibility.
