# AiExponent
> The EU AI Act's required artefacts, generated where engineers work.

## About
Open source tools that generate the EU AI Act's article-mapped compliance artefacts, with cross-mapping to NIST AI RMF and ISO/IEC 42001. Built by a practitioner, for practitioners.

- [Website](https://aiexponent.com): Product site and documentation.
- [GitHub](https://github.com/aiexponent): Source for every tool, Apache 2.0 licensed.
- [LinkedIn](https://linkedin.com/company/aiexponent): Company updates.
- [Contact](mailto:hello@aiexponent.com): General enquiries.

## Open Source Tools

- [License Compliance Checker](https://aiexponent.com/tools/license-compliance-checker): Framework-agnostic license compliance harness for AI systems and software dependencies. Install: `pip install license-compliance-checker`. Docs: https://aiexponent.com/docs/license-compliance-checker. Source: [github.com/aiexponent/license-compliance-checker](https://github.com/aiexponent/license-compliance-checker). EU AI Act Article 53, GPAI Compliance. Status: released. Language: python.
- [RiskForge](https://aiexponent.com/tools/riskforge): EU AI Act Article 9 Risk Management System builder: 30 minutes to a signed, audit-trailed Risk Management File. Install: `pip install riskforge`. Docs: https://aiexponent.com/docs/riskforge. Source: [github.com/aiexponent/riskforge](https://github.com/aiexponent/riskforge). EU AI Act Article 9, Risk Management. Status: released. Language: python.
- [Agentic Document Analyser](https://aiexponent.com/tools/agentic-document-analyser): VLM-powered document-to-structured-JSON pipeline for compliance evidence processing. Install: `docker compose up`. Docs: https://aiexponent.com/docs/agentic-document-analyser. Source: [github.com/aiexponent/agentic-document-analyser](https://github.com/aiexponent/agentic-document-analyser). EU AI Act Articles 11 + 19, Evidence Processing. Status: alpha. Language: docker.
- [RAG Benchmarking](https://aiexponent.com/tools/rag-benchmarking): Framework-agnostic evaluation harness for RAG and agentic AI systems. Install: `pip install rag-benchmarking`. Docs: https://aiexponent.com/docs/rag-benchmarking. Source: [github.com/aiexponent/rag-benchmarking](https://github.com/aiexponent/rag-benchmarking). EU AI Act Article 15, Accuracy Requirements. Status: released. Language: python.
- [LitmusAI](https://aiexponent.com/tools/litmusai): Free, deterministic CLI screener for the eight prohibited AI practices in EU AI Act Article 5. Install: `pip install litmus-screener`. Docs: https://aiexponent.com/docs/litmusai. Source: [github.com/aiexponent/litmusai](https://github.com/aiexponent/litmusai). EU AI Act Article 5, Prohibited Practices. Status: released. Language: python.

## Docs as Markdown
Every docs page and EU AI Act article page has a Markdown version at the same path plus ".md". Coding agents: start with [agents.md](https://aiexponent.com/agents.md).

- [Documentation](https://aiexponent.com/docs.md)
- [Install the tools](https://aiexponent.com/docs/install.md)
- [How the tools connect](https://aiexponent.com/docs/connect.md)
- [Roadmap](https://aiexponent.com/docs/roadmap.md)
- [License Compliance Checker](https://aiexponent.com/docs/license-compliance-checker.md)
- [RiskForge](https://aiexponent.com/docs/riskforge.md)
- [Agentic Document Analyser](https://aiexponent.com/docs/agentic-document-analyser.md)
- [RAG Benchmarking](https://aiexponent.com/docs/rag-benchmarking.md)
- [LitmusAI](https://aiexponent.com/docs/litmusai.md)
- [EU AI Act Article 5: Prohibited AI practices](https://aiexponent.com/eu-ai-act/article-5.md)
- [EU AI Act Article 9: Risk management system](https://aiexponent.com/eu-ai-act/article-9.md)
- [EU AI Act Article 10: Data and data governance](https://aiexponent.com/eu-ai-act/article-10.md)
- [EU AI Act Article 11: Technical documentation](https://aiexponent.com/eu-ai-act/article-11.md)
- [EU AI Act Article 13: Transparency and provision of information to deployers](https://aiexponent.com/eu-ai-act/article-13.md)
- [EU AI Act Article 15: Accuracy, robustness, and cybersecurity](https://aiexponent.com/eu-ai-act/article-15.md)
- [EU AI Act Article 19: Automatically generated logs](https://aiexponent.com/eu-ai-act/article-19.md)
- [EU AI Act Article 53: Obligations for providers of general-purpose AI models](https://aiexponent.com/eu-ai-act/article-53.md)

## EU AI Act Tool Mapping

| Tool | Article | Label | Description |
|------|---------|-------|-------------|
| License Compliance Checker | Article 53 | GPAI Compliance | Generates audit evidence supporting EU AI Act Article 53 documentation obligations: evaluates model card completeness, license compliance, and training data risk for AI components in your stack. |
| RiskForge | Article 9 | Risk Management | Produces structured Article 9 Risk Management Files for high-risk AI systems suitable for inclusion in your Annex IV technical documentation pack. Organises the assessment across 8 risk dimensions derived across Articles 9, 10, 13, 14 and 15 (health & safety, fundamental rights, discrimination, privacy, transparency, human oversight, robustness, and data governance), with cross-maps to NIST AI RMF and ISO/IEC 42001. Not a substitute for notified-body conformity assessment. |
| Agentic Document Analyser | Articles 11 + 19 | Evidence Processing | Structures unstructured compliance documents into machine-readable JSON for Article 11 technical documentation packages and Article 19 automatically-generated-log preservation workflows. Does not implement Article 9 risk management; use RiskForge for that. |
| RAG Benchmarking | Article 15 | Accuracy Requirements | Provides systematic accuracy testing and documentation for high-risk AI systems under Article 15. |
| LitmusAI | Article 5 | Prohibited Practices | Screens AI systems against the eight prohibited-practice categories of EU AI Act Article 5(1)(a)–(h). Conservative-by-default verdicts; UNREVIEWED reference ruleset (no external lawyer review yet); BYO signed-ruleset path for customers who need lawyer-reviewed output today. Article 5 has been applicable since 2 February 2025 (Art. 113(a)); sanctionable since 2 August 2025 (Art. 113(b)). |

## Not Yet Released
None of the following can be installed today. Full list: https://aiexponent.com/docs/roadmap

### Sigil
- [Release updates](https://aiexponent.com/contact?topic=sigil-updates): Free, open-source attestation for AI agent oversight. Not yet released.
- Status: Not released. v0.1 has no date yet. Free and Apache 2.0 when it ships. It will read and attest only, and will never intercept, block or proxy an agent action. v0.1 will cover EU AI Act Articles 14 and 26 plus Singapore IMDA; CBUAE, DIFC, SDAIA and the draft Saudi NCA AICG-1:2026 are targeted by v1.0.
- Details: Sigil will read the audit and trace exports your gateway, observability stack and identity provider already produce, then emit the evidence an examiner asks for: an Agent Register and an Oversight Log as hash-chained JSON and PDF. v0.1 will cover EU AI Act Articles 14 and 26 plus Singapore IMDA. CBUAE, DIFC, SDAIA and the draft Saudi NCA AICG-1:2026 are targeted by v1.0, alongside a signed kill-switch drill record. It will never intercept, block or proxy an agent action. Apache 2.0, runs locally, no telemetry. v0.1 has not been released and has no date yet.
- EU AI Act: Articles 14, 26, Agent Oversight Attestation
- Pricing: The tool is free and always will be. Everything that produces or verifies the artifact, including every regulator mapping, is free forever. A paid layer for portfolio views, examiner export packs, retention, SSO and on-prem packaging is milestone-gated and unpriced.

### Also committed
- OrgLiterate (Article 4): Sequenced after Sigil v0.1. A date will be published when one is set.

### Built on request, no schedule
TraceForge, TransparencyDeck, RMFMapper, ISOEvidence, HealthAI-Comply.

## Contact
- [General enquiries](mailto:hello@aiexponent.com): hello@aiexponent.com
- [Security](mailto:security@aiexponent.com): security@aiexponent.com
- [Website](https://aiexponent.com): https://aiexponent.com