Resources

Start with the question, then choose the source

Use an official source to establish the rule, a dated explanation to understand its context, and a working record to document your decision. This directory is selective; inclusion is not an endorsement of a provider or a claim that a tool satisfies a legal requirement.

For detailed legal citations and qualifications, follow the source guide and the footnotes in the relevant chapter.

Official source collections

• European Union: European Commission AI Act policy hub for implementation material; EUR-Lex for the enacted text and amendments. Start with the handbook’s EU AI Act chapter.
• United States: NIST AI Risk Management Framework, White House presidential actions and the relevant agency’s official publications. The federal chapter distinguishes these different sources of authority.
• State and local law: use the legislature’s enacted text, the current codified law and the responsible regulator’s notices. The states and cities chapter links to the provisions discussed.
• International comparisons: the OECD AI Policy Observatory is a discovery aid. Verify legal status against each jurisdiction’s official text; see international governance.
• Standards: the ISO artificial intelligence committee and official standard catalog entries identify editions and scope. Full standards may require a licensed copy; the ISO/IEC chapter does not reproduce their requirements.

Practical evaluation resources

These projects can support a defined evaluation, but no single metric resolves fairness, security or legal applicability.

• Fairlearn — tools and documentation for examining fairness in machine-learning systems.
• AI Fairness 360 — an open-source toolkit for examining and mitigating unwanted bias.
• Microsoft Responsible AI Toolbox — tools for evaluating aspects of model behavior.
• NIST AI Resource Center — supporting material for AI risk-management work.

Before adopting a tool, check its maintenance status, license, supported model and data types, privacy implications and fit to the decision. Record the version and limitations alongside the results. See technical safety for the evaluation workflow.

Continue within Techné AI

• Understand a specific issue: the public briefings connect selected legal sources to employment, board and governance questions.
• Start an employment-AI inventory: the free AI Hiring Exposure Snapshot helps organize initial questions; it is not a legal conclusion.
• Use editable employment documents: TalentSight Intelligence is a separate paid library. Review the current product page for its contents, price and terms.
• Use editable board documents: BoardSight Intelligence is a separate paid library, with its own product information and checkout.
• Read the complete publication offline: download the handbook PDF. It carries the same edition date as the web publication.

The handbook is free to read without an account. Buying a separate library is not required to access these chapters. See Terms for the website’s terms and the separate paid-product policy.

Report an issue

Use the corrections process to report an outdated date, broken source, accessibility problem or substantive concern. Include the chapter URL, passage and a primary source where available. A report is not a promise of an individual legal or technical assessment.