What the major frameworks teach, what they leave out, and who wrote them.
The two dimensions ALT-TEXT added, power and political economy and alternatives and refusal, are the least-covered in the whole corpus, tied with labour. Who owns AI, whether it can be refused, and the people who build it are precisely what the mainstream frameworks pass over. Ethics scores highest; ownership scores lowest.
Sixteen dimensions in four groups. Each square is one score, 0 (absent) to 3 (central to the framework). Click a square for the reasoning and source; a column to sort; a framework name for its full profile. The two ALT-TEXT dimensions are marked in purple.
Power scores set against what is on record about authorship and funding. Industry involvement means a company or industry-funded coalition issued, drafted, funded or sat on the expert group. A small sample, and a pattern to investigate rather than proof of cause.
What changed as two frameworks moved from consultation draft to adopted text. The pattern is telling: labour and critical material tend to be the first to go.
The material on concentration, industry in AI education, data labour and environmental cost that the power dimension is built to capture. Figures are dated; projections are marked.
Where mainstream frameworks mention alternatives, they usually mean the choice not to use AI. These initiatives offer something else. Money pledged and money reported spent are shown separately.
ALT-TEXT proposes a competence domain on who owns AI, who profits and whose interests it serves. The Power and Alternatives columns above are the first attempt to measure it. These competences are proposals, not yet coded against any framework.
Thirteen coded frameworks. Breadth is the sum of the sixteen dimension scores out of 48. Red marks a framework with documented industry involvement. Filter, then click any card for its definition, structure, authorship and full scores.