I can’t predict where we’ll end up with AI attribution out in society; whether we’ll continue to have an appetite for responsible disclosure of leveraging AI-generated/assisted content, or whether it would just be better to assume everything has been touched by AI at some point or the other.
I am not aware of a standard disclosure type for adding to code/code repos (i.e. README) files, or labels, but here is what I am workshopping. What do you think?
## 🤖 AI Assistance Disclosure
This project was developed with the assistance of Generative AI. The core logic, architecture, and styling were produced through a collaborative process between the human author and AI models.
**AI Usage Details:**
* **Models Leveraged:** [e.g., Gemini 1.5 Flash, GPT-4o, Claude 3.5 Sonnet]
* **Services Used:** [e.g., Google AI Studio, GitHub Copilot, Cursor]
* **Scope of Assistance:** [e.g., Boilerplate generation, CSS debugging, WordPress API implementation, and ASCII art refinement]
* **Verification:** All AI-generated code was manually reviewed, tested, and modified by the author to ensure system stability and security.
---
/**
* -------------------------------------------------------------------------
* AI ASSISTANCE DISCLOSURE
* -------------------------------------------------------------------------
* Portions of this file were generated or optimized using AI assistance.
* * MODEL(S): [Insert Model Name]
* SERVICE: [Insert Service Name]
* DATE: [Insert Date]
* * NOTE: The logic in this file has been human-verified for production use.
* -------------------------------------------------------------------------
*/
Also, it seems we’re settling on the following icons to denote some sort of AI involvement/features.
| Character | Name | Appearance |
| ✦ | Black Four Pointed Star | The most accurate “solid” version of the Gemini. |
| ✧ | White Four Pointed Star | An outlined version, similar to the Gemini logo in dark mode. |
| ✨ | Sparkles | The emoji version (often used to denote AI features). |