Law / Frameworks / Watch mode
Watch mode: framework drafts and candidates
Governance frameworks we follow before we read them: drafts their publishers have released for comment, documents a publisher has announced with no draft yet, and final frameworks we plan to read next. None of them is read against the law yet. When one is, it gets pages of its own under Frameworks and leaves this list.
Drafts
Released for comment, so they can still change before the final edition, or announced by their publishers with no text yet.
| Draft | Stage | Comments | On LexLint | Checked |
|---|---|---|---|---|
| Cybersecurity Framework Profile for Artificial Intelligence (Cyber AI Profile)National Institute of Standards and Technology, NIST IR 8596 iprd | Initial Preliminary Draftreleased NIST says the comments will inform an initial public draft. | to | Watched, not yet read against the lawwould sit beside NIST CSF 2.0 | |
| NIST Privacy Framework 1.1National Institute of Standards and Technology, NIST CSWP 40 ipd | Initial Public Draftreleased | to | Watched, not yet read against the lawa new edition of NIST Privacy Framework | |
| Managing Misuse Risk for Dual-Use Foundation ModelsNational Institute of Standards and Technology, NIST AI 800-1 2pd | Second Public Draftreleased No later draft or final has been published since. | to | Watched, not yet read against the lawwould sit beside NIST AI RMF | |
| Guidance and Templates for Public-Facing AI DocumentationNational Institute of Standards and Technology, AI Standards Zero Draft, initial public draft | Initial Public Draftreleased NIST plans a revision, possibly final, and then to propose it to a standards body. | until | Watched, not yet read against the law | |
| Secure Software Development Framework (SSDF) Version 1.2National Institute of Standards and Technology, NIST SP 800-218 Rev. 1 ipd | Initial Public Draftreleased It revises version 1.1, NIST SP 800-218. | to | Watched, not yet read against the law | |
| AI Risk ManagementCEN and CENELEC, Joint Technical Committee 21, prEN 18228 | Public Hearing (Enquiry)No later stage was found on JTC 21's or CEN-CENELEC's pages. | until | Watched, not yet read against the law | |
| Cybersecurity Specifications for AI SystemsCEN and CENELEC, Joint Technical Committee 21, prEN 18282 | Public Hearing (Enquiry)No later stage was found on JTC 21's or CEN-CENELEC's pages. | until | Watched, not yet read against the law | |
| AI Trustworthiness Framework, Part 1: LoggingCEN and CENELEC, Joint Technical Committee 21, prEN 18229-1 | Public Hearing (Enquiry)JTC 21 lists parts 2 to 5 (transparency, human oversight, accuracy and robustness) with no stage. | until | Watched, not yet read against the law | |
| Quality and governance of datasets in AICEN and CENELEC, Joint Technical Committee 21, prEN 18284 | Submitted to enquiry in November 2025No close date or result was found on CEN-CENELEC's pages. | none stated | Watched, not yet read against the law | |
| AI Risk Management Framework, revised editionNational Institute of Standards and Technology, NIST AI 100-1, revision | Being revised; no draft published | none stated | Watching for a first drafta new edition of NIST AI RMF | |
| Control Overlays for Securing AI Systems (COSAiS)National Institute of Standards and Technology, SP 800-53 control overlays | Outline only; no draft publishedFeedback on an outline of the first overlay closed on . | none stated | Watching for a first draftwould sit beside NIST CSF 2.0 | |
| AI RMF Profile on Trustworthy AI in Critical InfrastructureNational Institute of Standards and Technology, Concept note | Concept note only; profile in developmentNIST released the concept note on . | none stated | Watching for a first draftwould sit beside NIST AI RMF |
Candidates
Final frameworks we plan to read against the law next, each under a license that lets us print it without an agreement.
| Framework | Edition | License | On LexLint | Checked |
|---|---|---|---|---|
| Code of Practice on marking and labelling of AI-generated contentEuropean Commission, AI Office, EU AI Act Article 50 code of practice | Finalpublished The Commission licenses its own content under CC BY 4.0 unless a document says otherwise. | CC BY 4.0 | Not yet read against the law | |
| OWASP AI Testing GuideOWASP Foundation, AI Testing Guide | Version 1published | CC BY-SA 4.0 | Not yet read against the law |
Already read against the law, while still drafts: NIST AI 200-2 Draft, MIT mitigations Draft. Each has pages of its own, and its publisher's page is read on the same schedule as the drafts above.
How this list is kept
Each entry is read by a person on its publisher's own page, and the day of that reading is the one in its row. Every week LexLint also reads each publisher's page, and a sign that a draft has moved on (a final edition, a new draft, a page that is gone) sends the entry back to a person. A draft read against the law gets pages of its own and leaves this list, and so does a candidate.
Every draft released, every final edition and every framework read against the law is dated on What changed in the frameworks.
Laws a government has announced and published no text for are on Watch mode: planned laws.