Quick triage
Run your system spec here:
- /eu-ai-act
What transparency means
Even when a system is not high-risk/prohibited, transparency duties may apply depending on:
- Human interaction where users might not realize it’s AI
- Synthetic or deepfake content generation
Mermaid source
flowchart LR\n I[Human interaction] --> D[Disclosure signal]\n G[Generated synthetic media] --> D\n D --> R[Audit output + links]
What to capture in your spec
- Whether users know they interact with AI
- Whether content is synthetic / deepfake-like
- Where the AI output is shown (UI surfaces)
Next guide
- EU AI Act FAQ → /policy-guides/eu-ai-act-faq