Watermarking
Embedding a mark into digital content to indicate AI generation or significant manipulation.
How the laws define this
Both sources agree that watermarking is a process or method of embedding information into digital media or content, such as images or videos, to indicate its origin, specifically regarding content that has been generated or manipulated by AI. The key divergence between the texts lies in the specificity of the embedded material and the degree of manipulation required. Under California Executive Order N-5-26 — Responsible Procurement and Deployment of Generative Artificial Intelligence, watermarking requires embedding a discernible mark or data, and it explicitly applies to content that was generated or significantly manipulated by AI. By contrast, in As Trump rolls back protections, Governor Newsom signs first-of-its-kind executive order to strengthen AI protections and responsible use, the term is defined more broadly as embedding identifying information, applying particularly to AI-generated or manipulated content without specifying that the mark must be discernible or that the manipulation must be significant.
Synthesised from the 2 statutory definitions below. Each one is quoted in full, with its source.
Definitions (2)
The process of embedding a discernible mark or data into digital content (e.g., images, video) to indicate its origin or that it was generated or significantly manipulated by AI.
A method of embedding identifying information into digital media, such as images or videos, to indicate its origin, particularly for AI-generated or manipulated content.