Six major frameworks.
All missing the same things.

I — The Moment

2 August 2026: the EU AI Act enters its main application and enforcement phase.

The EU Artificial Intelligence Act entered into force in August 2024 and is being implemented progressively. From 2 August 2026, the majority of its rules apply and enforcement begins for applicable provisions, including Article 50 transparency requirements — though the regime remains phased: Annex III high-risk AI rules apply from December 2027, and high-risk AI embedded in regulated products from August 2028. The AI Act has significant extraterritorial reach. In defined circumstances, obligations can apply to providers and deployers outside the EU, including where AI-system outputs are used in the Union.

But the Act was always designed to work alongside voluntary frameworks, not replace them. Article 95 explicitly creates space for providers to self-organise into codes of conduct, and calls on the Commission to encourage their development. The Charter Mark sits squarely in that space: a voluntary, human-validated complement to mandatory compliance — covering territory the Act's statutory provisions were never designed to reach.

II — The Consensus Seven

These seven principles form the common foundation. Most frameworks approach them differently.

In 2019, the European Commission's High-Level Expert Group on AI published Ethics Guidelines for Trustworthy AI — seven requirements that have become an influential reference point for global AI governance. The OECD Principles substantially overlap with them. NIST's AI Risk Management Framework maps to comparable territory. UNESCO's 2021 Recommendation on the Ethics of Artificial Intelligence builds on similar ground. Each framework organises and operationalises these principles differently, but the common territory is clear.

The Sacred Logic Human Values Framework begins with the same seven. Points 1–7 are aligned with the seven requirements identified by the EU High-Level Expert Group. Where an organisation can provide robust, independently verifiable evidence from applicable AI Act conformity assessments or equivalent assurance processes, Sacred Logic may recognise relevant evidence towards Points 1–7 through the Compliance Fast Track — subject to review against the Charter Mark criteria, which in some areas go further than statutory obligations.

The question is what happens after point seven.

Not one of the major frameworks requires that AI involvement in creative or intellectual outputs be disclosed, that attribution and moral rights be preserved, or that AI output not be misrepresented as unaided human work. These are not edge cases. They are the defining governance questions of the generative AI era.

IV — The Unfilled Gap

Authorship and culture: why they were left out, and why they can't stay out.

The major frameworks were built before generative AI dominated the conversation. The HLEG guidelines were finalised in April 2019, three and a half years before ChatGPT made the authorship question inescapable. The frameworks were not wrong to focus on the high-risk systems that already existed — they were responding to the world as it was.

The world has changed. AI systems now produce text, images, audio, and code at scale, in professional contexts, and in ways that are frequently indistinguishable from human output without explicit disclosure. The governance frameworks have not caught up.

The UNESCO Recommendation on the Ethics of Artificial Intelligence (2021) comes closest, noting that AI systems should protect "cultural diversity and the richness of pluralism." But it stops at aspiration — there is no operationalised requirement, no evidence standard, no audit mechanism. A named concern is not a certification criterion.

Points 8, 9, and 10 of the Human Values Framework fill this gap with evidence-backed, auditable requirements — a distinctive approach not generally treated as explicit, auditable certification criteria within the major frameworks compared above.