ThirdLaw

ThirdLaw builds a runtime enforcement and control layer for enterprise AI. Its platform captures prompts, responses and tool calls, evaluates them against enterprise policies in real time, and lets Security and IT teams respond to violations by blocking, redacting, rerouting or escalating the offending interaction. The system produces end-to-end incident traces and audit-ready evidence, positioning it at the intersection of AI safety, AI security and LLM observability.

The product, ThirdLaw Runtime Enforcement Platform, is aimed at organisations deploying AI agents at scale, particularly regulated enterprises. Its technical scope spans policy evaluation, data loss prevention and redaction, AI agent governance, and incident tracing. Enforcement latency is a core design constraint: evaluation and action on AI behaviour are performed in real time rather than as a post-hoc audit step, and coverage extends across prompts, responses and tool calls.

The founding team has experience at Splunk, VMware, MongoDB, Cloudera, Postman, Tintri, Vectara and Intel, spanning enterprise security, observability and AI product development. The company does not disclose its founding year, team size or funding. It operates in a category that Gartner has flagged as an emerging priority for organisations deploying AI at scale.