Clutch Security: AI Agents Have Surpassed Technical Staff in Enterprises and Are Doubling Rapidly
Clutch Security's research shows that AI agents have outnumbered technical staff in 84% of enterprises, with numbers doubling approximately every ten weeks.
The median number of non-human identities in use is 15, but most security teams are completely unaware of them.
This finding highlights event-driven enterprise security risks, strengthening market demand for AI agent governance tools, benefiting identity and visibility security targets, while pressuring enterprises with unmonitored IT environments.
Source: Public Information
ABAB AI Insight
Clutch Security reveals that the adoption speed of AI agents far exceeds procurement and security review processes by monitoring actual agent activities across millions of endpoints, focusing on the visibility gap of non-human identities and AI agents.
In terms of capital, enterprises are shifting from experimental deployments to large-scale agent usage, prioritizing resources over business efficiency rather than security governance, motivated by the desire to capture automation benefits, leading to credential and identity loss of control.
Similar cases can be seen in the early spread of shadow IT in cloud computing and API keys, as well as the current rise of non-human identity management; enterprises are currently in a phase of AI agent explosion but lagging in governance, with the industry's position shifting from tool experimentation to a necessity for visible infrastructure risk.
This essentially reflects a governance lag in technological substitution. As the number of agents grows exponentially, security visibility becomes a bottleneck, prompting capital to accelerate towards solutions that can provide agent discovery and credential control.
ABAB News · Law of Cognition
- Adoption speed far exceeds security review
- Invisible agents constitute the greatest risk
- Governance requires achieving visibility first.