Adobe has launched a new web tool that enables creators to attach content credentials to their images, giving them more control over how their work is used to train AI models. The tool, called Adobe Content Authenticity App, allows users to add their credentials, including name and social media accounts, to a file, and signal to AI companies whether they can use the image for model training.

The move is part of Adobe's effort to create a standard for content authenticity, similar to the robots.txt file that websites use to indicate which crawlers are not allowed. The company is partnering with LinkedIn to verify the identity of creators, and users can also attach their Instagram or X profiles to an image. However, there is no integration with verification of these platforms.

The Adobe Content Authenticity App is built on the Coalition for Content Provenance and Authenticity (C2PA) standard, which aims to provide a way to identify the authenticity and ownership of digital content. The app uses a mix of digital fingerprinting, open source watermarking, and crypto metadata to embed metadata in various pixels of an image, making it difficult to remove or modify.

While Adobe's intentions are commendable, the success of the initiative depends on AI companies adopting and respecting the standard. The company is in talks with top AI model developers, but has not yet signed any agreements. If AI companies ignore the signals, the tool will be ineffective.

This is not the first time the issue of AI model training has sparked controversy. Last year, Meta's implementation of labels to auto-tag images on its platform caused an uproar among photographers, who complained about their edited images being tagged with a "Made with AI" label. Meta later changed the label to "AI info."

Andy Parson, Senior Director of the Content Authenticity Initiative at Adobe, emphasized that the company wants to give creators a way to signal their intent about AI platforms. "Content creators want a simple way to indicate that they don't want their content to be used for gen AI training. We have heard from small creators and agencies that they want more control over their creations," Parson said.

Adobe is also releasing a Chrome extension that allows users to identify images with content credentials. The extension will display a small "CR" symbol on an image if it has content credentials attached. The company plans to add support for video and audio files in the future.

The debate around AI and art is ongoing, with questions about ownership and attribution. Parson believes that content credentials could be an important marker for ownership, allowing artists and creators to sign their work and claim attribution. However, he emphasized that C2PA does not aim to opine or direct what constitutes art.

As the use of AI models becomes more widespread, the need for clear guidelines and standards around content authenticity and ownership will only increase. Adobe's initiative is a step in the right direction, but its success will depend on the willingness of AI companies to adopt and respect the standard.