AIActAudit

AIActAudit — Deep Dive

The Problem The EU AI Act entered full enforcement in 2026. Every company deploying an AI system touching EU users is now legally obligated to:

Classify their system under the Act's risk tiers Fulfill tier-specific obligations (documentation, conformity assessments, human oversight, etc.) Register high-risk systems in the EU AI Act database Face fines up to €35M or 7% of global turnover for violations There is no accessible, developer-friendly tooling for this. AIActAudit fills that gap.

The system takes a natural-language description of an AI deployment (or a technical spec doc), runs it through a multi-stage agentic pipeline, and produces a structured compliance report that a non-lawyer can act on.

Built With

• auth
• claude
• langgraph
• pydantic
• rag
• react
• vectordb