R0029/2026-03-27/Q001/SRC03/E01¶
AIA (Artificial Intelligence Attribution) icon system for legal and ethical disclosure
URL: https://scholarlycommons.law.northwestern.edu/njtip/vol22/iss1/1/
Extract¶
Avery, Abril, and del Riego propose the AIA (Artificial Intelligence Attribution) system — a set of visual icons designed to communicate the level and nature of AI involvement in text generation. The system defines graduated levels of AI involvement:
- Generated: Content fully created by AI
- Edited: Content generated by AI but revised by humans
- Suggested: AI provided suggestions that humans incorporated
The system is modeled on existing visual disclosure systems (similar to Creative Commons icons) and is designed to be "seamlessly" integrated into published works. The proposal addresses both legal disclosure requirements and ethical norms around transparency.
The article is framed as addressing "a gap in standardized disclosure systems for AI contributions in text generation" and draws inspiration from how CRediT taxonomy handles contributor attribution in academic publishing.
Relevance to Hypotheses¶
| Hypothesis | Relationship | Strength |
|---|---|---|
| H1 | Supports | A formal, published framework exists with specific structural components |
| H2 | Contradicts | The graduated icon system is distinctly more structured than binary disclosure |
| H3 | Supports | Published as a law review proposal ("towards a system") — not yet adopted or standardized |
Context¶
The proposal comes from the legal/IP domain rather than the computer science or publishing domains. This cross-disciplinary origin suggests the need for AI attribution frameworks is recognized across fields, but also means adoption would require buy-in from multiple stakeholder communities.