The initiative to develop systems for "watermarking AI-generated media" for IP protection and trust assurance is, in some ways, related to the broader effort to create AI that embeds targeted subliminal messages for use in influence operations and advertising.
