AI DisclosureGENAIS level BContent developed by shared human and AI input. Concept developed by human but detailed by AI.Drafted with GenAI, then rewritten by me.How this site discloses AI use →
I want to practice and promote safe use of Generative AI systems. Part of that is normalising the practice of marking work assisted by GenAI systems to let audiences know what they’re consuming. This might seem odd given we’ve used computer tooling and Artificial Intelligence for years without worrying about disclosure, however I believe it’s a useful and necessary step to combat misinformation and to differentiate between GenAI being used as the right tool for the job, and GenAI being used to mislead.
I’ve adopted the proposed GENAIS framework for disclosure as I feel it gives reasonable granularity on GenAI usage. The GENAIS symbols broadly describe whether work was human led or GenAI led. This website has a broad ‘site wide’ disclosure at the bottom of every page, and individual assets and texts are marked with GENAIS symbols to show how they were created. At any time you can hover or click on a GENAIS symbol for details.
The site itself
This site was specified and designed by me and built by GenAI.
- I chose layouts, styles, colours, fonts and used GenAI to apply those.
- I described the AI-native system I wanted this website to be built on and had GenAI detail it.
- I transcribed my work history, project experiences etc and used GenAI to interview me to flesh out gaps.
- I’ve used GenAI to review and occasionally adjust framing for my work experience, portfolio, and projects.
- While my writing is my own, the supporting images are individually marked to show their level of GenAI involvement.
The scale
Broadly the GENAIS symbols describe the work split between human and GenAI. The meanings are quoted from the paper.
| A | Content human developed but structure and/or format enhanced or validated by AI. | |
|---|---|---|
| B | Content developed by shared human and AI input. Concept developed by human but detailed by AI. | |
| C | Content largely developed by AI but enhanced with human input or final edits. | |
| D | Content entirely generated by AI prompt. May include images, video, text. |
How I choose a level
- I rate what you’re looking at, not what it’s about. A video I recorded of an AI system gets no symbol. It’s about AI, not made by it.
- The footer’s symbol covers the site itself. That’s the code, layout and chrome. Anything with its own mark is rated on its own.
- There’s no “no AI” symbol. Work I made myself gets no symbol, just a short credit line, and third-party work is credited to its source.
- Where it came from outranks how much I polished it. A generated image I’ve recomposed, lettered and animated is still C.
- When I’m unsure between two levels, I pick the higher one.
- If a symbol and the text beside it seem to disagree, the text explains why. The symbol rates the asset. The text can describe how it was made.
The chatbot
The Lia Chatbot was iteratively designed by me and built by GenAI. It’s a test platform for me. It uses a combination of Conversational AI systems: you’ll see a mixture of responses written in advance with GenAI’s help, generated in the moment based on RAG, or generated in the moment based on persona design. Each response is marked with its symbol: B for prepared replies, D for generated ones.
Retroactive labels
Disclosures were added in September 2026, including to works published before then.
Source
Glassey, S. (2025). Standardising AI Disclosure in Academia: The GENAIS Symbology Approach. Preprints. doi:10.20944/preprints202510.2183.v1. The paper is CC BY 4.0; the symbols are reproduced from it with attribution. GENAIS™ is the author’s trademark. The official font is at tinyurl.com/genais-font.