Users of Adobe apps, including Photoshop and Substance Painter, were recently notified of changes to Adobe’s Terms of Service that have caused confusion and outrage. The updated terms, which came into effect on February 17, 2024, state that Adobe may access user content, including unpublished and in-progress projects, through both manual and automated methods for content review. This has raised concerns among users that their creative work could be used as training data for Adobe’s AI models.
The language used in the updated terms, specifically mentioning “automated systems” and using “machine learning to improve our Services and Software,” has sparked fears that users’ content will be used without credit or compensation. In addition to the potential lack of recognition for their work, users are also worried about the privacy implications, particularly for those working with confidential information. The idea that Adobe would have full access to anything created using Photoshop, including NDA work, is a cause for concern.
While Adobe has clarified that it does not use content stored locally on users’ devices, only content stored in the Creative Cloud, there are still concerns about the use of public content. Adobe has been using public content, such as contributions to Adobe Stock and submissions for tutorials in Lightroom, to train its AI algorithms and improve its products and services. However, this practice raises questions about copyright infringement and unauthorized works seeping into model training. It has also raised doubts about the ethical use of AI features and eroded customer trust in Adobe’s practices.
It’s worth noting that the recent policy change by Adobe has not definitively exposed users to privacy invasions. The updated terms mention the need to access user content for purposes such as detecting fraud, addressing security issues, and enforcing its terms of service. However, the vague language used, coupled with concerns about machine learning and data privacy, has understandably alarmed users who value their privacy.
Adobe has not explicitly mentioned its AI model Firefly or its use of AI training data in the updated terms of service. Firefly, which generates images and powers other AI features, was trained on public domain data, including AI-generated images from a competitor. This has further added to the complexity of gathering training data and ensuring copyrighted content is not used without permission.
In response to the concerns raised by users, Mashable has reached out to Adobe for clarification. The outcome of this inquiry will shed light on the extent to which user privacy may be compromised and provide clarity on Adobe’s policies regarding AI training data.
In conclusion, the changes to Adobe’s terms of service have sparked confusion and outrage among users. The language used in the updated terms, coupled with concerns about privacy and the use of AI training data, has raised valid concerns about the protection of users’ creative work. Clarification from Adobe is needed to address these concerns and rebuild trust with its user base.
