Arcjet Launches Runtime Security Product for AI Agents
Security teams can review AI agent activity and govern individual workflow steps while they run.
- By Danielle Naidu
- Sep 18, 2026
A new Arcjet product gives security teams a way to monitor and control AI agent behavior during active workflows. Announced Sept. 17, the runtime security offering also creates execution records for subsequent security and compliance reviews.
Applications receive policy decisions from Arcjet before protected actions proceed. Developers can use those decisions to block an operation or seek human approval. Available controls include usage limits, automation detection, prompt injection screening and measures to prevent or redact exposure of personally identifiable and other sensitive information.
Arcjet Guards evaluates deterministic rules through Rego and Open Policy Agent across interactions with large language models, tools, databases and application programming interfaces. Existing OpenTelemetry event feeds can supply activity data without installing another software agent or modifying application code, the company said. Information from Claude environments can also enter through the Claude Compliance API.
Organizations can retain execution records through Arcjet’s cloud, a single-tenant environment or a private VPC. The records can include identities, inputs, actions and policy outcomes. Arcjet does not categorize the product as a security information and event management system, but logs can be forwarded to existing detection tools. Native framework support includes the Claude Agents SDK, OpenAI Agents SDK and LangChain.
About the Author
Danielle Naidu is assistant editor for Security Today, Campus Security Today, Occupational Health & Safety and Environmental Protection Online.