TrendAI adds Claude session visibility for security teams

TrendAI has expanded its Vision One platform to include Claude Enterprise compliance records for Claude Code and Cowork sessions, adding AI agent session visibility to its existing security monitoring system.

Security teams using Vision One can now view prompts, responses, and tool calls from Claude Code and Cowork alongside endpoint, identity, cloud, and network telemetry. The addition builds on an existing integration with Anthropic's Claude Compliance API and is intended to give investigators a fuller view of how AI agents are used inside organisations.

The expansion comes as more businesses allow AI agents to write code, call tools, and access internal systems on behalf of staff. That shift has created a new monitoring challenge for security teams: they may know the tools are active but still lack visibility into what users asked them to do, what actions they took, and whether those actions aligned with internal policy.

Claude's Compliance API returns session records for Claude Code and Cowork, including prompts, responses, and tool calls linked to verified users. Vision One brings those records into the same environment used for broader security investigations, allowing analysts to correlate AI activity with other signals instead of treating it as a separate source of information.

Teams can investigate whether an agent made a tool call outside expected limits, tie a session to a specific user during an inquiry, and review whether a workflow followed required internal processes. They can also retain an auditable record of AI session activity for compliance purposes.

The update expands TrendAI's existing logging coverage. Vision One already ingests activity logs from Claude Enterprise, including user logins, administrator actions, and configuration changes, as well as logs from Claude Platform covering administrator, system, and resource events.

This latest addition focuses on session transcripts from tools used by engineering and knowledge-work teams. By bringing those transcripts into a central investigation platform, TrendAI aims to address a gap that has become more visible as AI agents take on routine operational work inside businesses.

The announcement is likely to resonate with organisations in Australia and New Zealand, where concerns about agentic AI security have grown after several widely reported incidents. Businesses have been under pressure to show that AI tools can be monitored and reviewed with the same discipline applied to other digital systems.

The integration is available now for Vision One customers using Claude Enterprise. TrendAI positions the feature as part of a broader effort to give security and governance teams more direct oversight of AI activity inside their organisations.

Rachel Jin, Chief Platform and Business Officer and Head of TrendAI, said visibility is now essential as AI tools become embedded in day-to-day work. "We understand security teams need transparency to validate AI agent activity against company policy and integrate that intelligence with their existing endpoint, identity, network and cloud signals," Jin said.

She added: "Coding and knowledge-work agents are now doing real work inside the business, and security teams should be able to see and investigate that work the same way they do everything else. This is about making informed decisions and maintaining control as these tools scale. Organisations can now move faster with AI while keeping proper oversight in place."