Recent changes in Adobe’s Terms of Use have made many users worried about the possibility of their data being used to train Adobe Firefly, Adobe’s new AI model. Concerns have been raised on how Adobe can use the creative content created by the users to build better AI algorithms.
Also Read: Adobe Summit 2024 – Unlocking the Potential of Personalization at Scale
Adobe, however, has issued a statement to weather the storm. It told users that the policy was aimed at curbing harmful content and that it would not use customer data to train its AI models as many had speculated.
Backlash erupts over ambiguous language
The backlash started when Adobe informed its users of changes to its Terms of Use. A sentence in Section 2.2 caught users’ attention: “Our automated systems may analyze your Content and Creative Cloud Customer Fonts…using techniques such as machine learning in order to improve our Services and Software and the user experience.” This made users assume that their confidential and unpublished work could be used in training the AI without their consent.
Also Read: Adobe Brings Superpower Generative AI Tools To Photoshop
In a blog post, Adobe stated that this is not a new policy. The policy is employed to delete abusive content, such as child sexual abuse material, as well as content violations, such as spam and phishing. Thus, the blog stated that more human moderation has been integrated into the content review process and not the automated one.
“Given the explosion of Generative AI and our commitment to responsible innovation, we have added more human moderation to our content submissions review processes.”
Adobe
Adobe explains policy changes
Adobe also gave details of what had been changed in the Terms of Use document. The revisions include replacing the phrase ‘will only’ with ‘may access, view, or listen to your Content’ and the inclusion of the phrase ‘through both automated and manual methods, but only’ in limited ways.’
The paragraph now also points to Section 4. 1 regarding prohibited or restricted content and includes the term ‘manual review’ to specify the content moderation processes. The language was also updated from “child ‘pornography'” to “child ‘sexual abuse material.'”
Also Read: Adobe Invests in Video Content to Boost AI Capabilities
Adobe clarified that these changes do not represent a shift in the company’s privacy policy. This assurance may help allay the fears of users who rely on Adobe for their creative projects. Adobe reaffirmed its position that the content created by individuals using Adobe software is owned by the individual. Adobe claims that the Firefly AI models are trained on licensed content such as Adobe Stock and public domain content and not user-generated content.
“Our commitments to our customers have not changed.”
Adobe
The company admitted that the response to the ambiguous update was a reflection of the anxiety and suspicion among creatives as AI generating capabilities advance. Similar concerns have recently been seen regarding Slack’s revised AI guidelines. The company was accused of using customers’ information to train its machine learning algorithms, the business messaging app.
Cryptopolitan Reporting by Brenda Kanana
