Adobe updated its Terms of Use after an AI-training backlash to state it won't train generative AI on customer content

What happened

In early June 2024, Adobe pushed a Creative Cloud Terms of Use update that Photoshop and other app users were prompted to accept. As Malwarebytes Labs reported, users balked at licensing language granting Adobe “a non-exclusive, worldwide, royalty-free sublicensable, license, to use, reproduce, publicly display, distribute, modify, create derivative works based on, publicly perform, and translate the Content” (source). Many creators read this as permission to train generative AI on their uploaded work, and the backlash spread (source).

This is a Terms-of-Service change made under pressure, not a technical deployment. On June 10, 2024, Adobe (Scott Belsky and Dana Rao) published “Updating Adobe’s Terms of Use,” stating plainly: “We don’t train generative AI on customer content. We are adding this statement to our Terms of Use to reassure people” (source). In the same post Adobe said a customer’s content “will never be used to train any generative AI tool,” and that “Adobe Firefly is only trained on a dataset of licensed content” such as Adobe Stock and public-domain material (source).

The outcome

Adobe said it would roll out the updated terms by June 18, 2024; Malwarebytes Labs reported that Adobe expected to “roll out new terms of service on June 18” (source). After the rollout, No Film School reported that the updated Terms of Service state Adobe “will not use your Local or Cloud Content to train generative AI” (source). Adobe’s position is that it never trained generative AI on customer content; what changed after the backlash was the addition of explicit Terms-of-Use language (source).

Weakest load-bearing source. The primary document here, Adobe’s own June 10 blog (Tier 1), carries the commitment, but the exact final wording of the rolled-out Terms of Service (“will not use your Local or Cloud Content to train generative AI”) and the June 18 rollout date come from independent newsrooms (Malwarebytes Labs and No Film School, both Tier 2), not from a byte-tied capture of the live updated Terms of Use page itself. Until the amended ToS page is captured and bound to that clause, the final legal wording rests on secondary reporting, not the primary legal instrument.

Note also a conflation guard: a separate December 2025 proposed class action alleging misuse of authors’ work in AI training is a later, distinct dispute and must not be merged with this June-2024 terms-change event.

How this was verified

• Method: Every quoted figure and clause was bound verbatim to a Wayback capture of the cited source. Adobe’s first-party blog (capture 20240610215004), Malwarebytes Labs (capture 20240612184251), and No Film School (capture 20240619202926) were each decoded and searched for the exact quoted strings; all bind. First-party origin (Adobe) is not counted toward independent corroboration; the two Tier-2 newsrooms are.
• Date verified: 2026-08-09, re-verified 2026-08-25 (all quoted clauses and the June 18 rollout date still bind on the live Adobe blog and the Malwarebytes report).
• What would further harden this: the live or archived amended Terms of Use page byte-tied to the reported clause (“will not use your Local or Cloud Content to train generative AI”), which would move the exact final legal wording from secondary reporting onto the primary legal instrument. Green never depends on Adobe confirming its own numbers as testimonial; it depends on the public record.

Sources

1. Adobe (Scott Belsky and Dana Rao) · “Updating Adobe’s Terms of Use” · 2024-06-10 · Tier 1 (first-party primary) · https://blog.adobe.com/en/publish/2024/06/10/updating-adobes-terms-of-use
2. Malwarebytes Labs (Pieter Arntz) · “No AI training in newly distrusted terms of service, Adobe says” · 2024-06-12 · Tier 2 (independent newsroom) · https://www.malwarebytes.com/blog/news/2024/06/no-ai-training-in-newly-distrusted-terms-of-service-adobe-says
3. No Film School (Jourdan Aldredge) · “Adobe’s Stance on AI” · 2024-06-19 · Tier 2 (independent newsroom) · https://nofilmschool.com/adobe-stance-on-ai

Related case files

• WeTransfer removed an AI-training clause from its terms after backlash (2025). The closest sibling: a file-transfer service walking back ToS language read as AI-training consent.
• Mozilla rewrote Firefox’s Terms of Use after backlash over data-for-AI wording (2025). Another vendor amending ToS language when users read it as AI-training permission.
• Dropbox clarified its third-party-AI (OpenAI) toggle after backlash (2023). A SaaS provider adding an explicit “not used to train models” clarification under pressure.
• Slack updated its AI privacy principles after user backlash over ML training (2024). The same pattern in a workplace tool: opt-out and policy language added post-backlash.