Traceforce

Traceforce is an AI security and control platform that runs on employee devices rather than at the network edge. Its premise is that AI usage has moved beyond the browser, so it monitors and controls browser AI tools, desktop apps such as ChatGPT, Claude and Cursor, and command line agents such as Claude Code, along with the MCP servers and skills those agents call. The agent installed on each device builds an inventory of the AI apps and MCP connectors in use, tracks tool calls, and blocks dangerous actions before they run. A feature called TraceGraph reconstructs incidents, showing prompts, tool calls, reasoning and outcomes instead of only process logs. The platform can also patch AI packages on the device under policy with audit logs. It deploys through existing MDM tools including Jamf, JumpCloud and NinjaOne, and the vendor describes onboarding as a self serve process of about 15 minutes. An open source MCP scanner is also offered. It is built for security and IT teams that need visibility into AI tools employees adopt on their own, including detection of sensitive data leaking into assistants like ChatGPT and Copilot. The pricing page lists a free trial at $0 for up to 10 devices for up to 30 days with no credit card, a Pro plan priced per device at $15 with unlimited devices, and a custom Enterprise plan that adds single sign on, multi tenant support and dedicated support.