R0029/2026-03-27/Q001/H3¶
Statement¶
Some structured proposals have been made but the field is pre-standardization, with no dominant framework adopted at scale.
Status¶
Current: Supported
The evidence consistently shows an emerging but immature landscape. Multiple independent teams have proposed structured attribution approaches, but none has achieved adoption by a standards body, widespread institutional uptake, or consensus status. The IBM toolkit is self-described as experimental. The AIA system is a law review proposal. The CRediT taxonomy has not been extended. Publisher policies remain at the disclosure level, not the structured attribution level.
Supporting Evidence¶
| Evidence | Summary |
|---|---|
| SRC01-E01 | CHI 2025 research calls for "new attribution approaches" — implying current ones are insufficient |
| SRC02-E01 | IBM describes their toolkit as "a first pass at formulating what a voluntary reporting standard might look like" |
| SRC03-E01 | AIA system proposed in an academic journal — a proposal, not an adopted standard |
| SRC04-E01 | CRediT, the closest existing standard, does not yet cover AI contributions |
Contradicting Evidence¶
No evidence contradicts this hypothesis.
Reasoning¶
Every piece of evidence points to the same conclusion: the field is in an active proposal phase but pre-standardization. Multiple independent efforts exist (IBM, Northwestern/Miami law, CHI research community), demonstrating recognized need. But no standards body has acted, no framework has achieved broad adoption, and the proposals themselves vary significantly in approach (toolkit vs. icons vs. taxonomy extension). This is characteristic of an emerging field that has not yet converged.
Relationship to Other Hypotheses¶
H3 is the most accurate characterization. It subsumes H1's correct observation (proposals exist) while correcting its overstatement (they are not established). It explains why H2 is wrong (structured work exists) while acknowledging the practical reality that most institutions still operate at the disclosure level.