Changelog — AI Liability
Frozen Baseline discipline. Material changes only.
Version 1.0
Published: 2026-04-14
- Initial structural baseline publication.
- Definition of AI liability established as a framework for allocating responsibility, accountability, and risk across artificial intelligence systems.
- Scope boundaries formalized (included / excluded).
- Structural phase model introduced (system creation → deployment context → operational behavior → responsibility attribution).
- Institutional reference sources documented.
- Mandatory legal, privacy, terms, and license pages implemented.
- 2026-05-17 — Added root-level llms.txt reference file.
- 2026-07-22 — Updated llms.txt to LLMS Baseline v2.0. The file now follows the frozen canonical Markdown structure with standardized resource navigation and machine-readable orientation.
- 2026-09-25 — Corrected
/license/, which had inadvertently duplicated the Privacy page, including its title and canonical reference. Restored the page to the applicable License v3.0 baseline. No change to the domain’s semantic definition, scope, or structural model.
Status: Frozen Baseline
Change Policy
Changes are recorded only if:
- The definitional boundary materially evolves.
- A structural clarification alters conceptual framing.
- Primary institutional reference frameworks materially change.
- A compliance-relevant correction is required for accuracy.
Minor editorial, stylistic, or formatting corrections are not versioned.
Versioning Logic
Version increments follow semantic clarity rules:
- Major increment — Definitional or structural shift.
- Minor increment — Material clarification without conceptual shift.
- No increment — Editorial refinement only.