Adobe's Latest Effort to Address Terms of Service Controversy: 'We Don't Use Customer Content for AI Training'

Today in “Complaints Get Results”: Adobe is responding to customer backlash over its recently updated Terms of Service (ToS). After users voiced concerns and threatened to cancel Adobe Creative Cloud subscriptions, the company announced it will provide a revised ToS next week. This update aims to clarify and address community worries.

In a blog post, Adobe reassured users with the statement: “We don’t train generative AI on customer content.” The clarification confirms that Adobe Firefly is trained solely on licensed content from Adobe Stock and public domain works where copyright has expired.

Key Changes and Clarifications:

Adobe's upcoming ToS update will emphasize several crucial points to reassure its extensive user base:

- Content Ownership: Customers retain full ownership of their content; Adobe confirms it will never utilize customer content to train generative AI tools.

- Generative AI Practices: Adobe Firefly is trained exclusively on licensed datasets and public domain content. Customer content is not utilized for training purposes.

- Product Improvement Programs: Users have the option to opt out of programs that employ data to enhance user experiences through machine learning, exclusively for non-generative AI applications.

- License Clarifications: Adobe will provide clearer license explanations, ensuring ownership remains with the customer while outlining specific permitted activities.

- Content Scanning: Adobe does not scan content on users' local devices. For content uploaded to its servers, the company scans for child sexual abuse material (CSAM) and other illegal content, with human reviews occurring only when flagged or requested by users.

Response to Public Backlash:

The backlash originated from a clause in the Terms of Use allowing Adobe to access and analyze user content via automated and manual methods for service improvement. Many users viewed this as a privacy concern, particularly regarding sensitive or client-protected information.

Adobe clarified that such measures are necessary for innovative features like Photoshop Neural Filters and Adobe Express’s Remove Background tool, while also assuring users that it cannot access locally stored content.

Renewed Commitment to Customer Trust:

Scott Belsky, Chief Product Officer, and Dana Rao, Chief Trust Officer, stressed Adobe's dedication to transparency and responsible innovation. “We’ve never trained generative AI on customer content, taken ownership of a customer’s work, or accessed customer content beyond legal requirements,” the company stated.

Adobe acknowledged the need to update its Terms of Use to align with the changing technological landscape and content creators' needs. The company pledged to maintain an ongoing dialogue with users and incorporate their feedback into future updates.

Future Initiatives:

In addition to the revised ToS, Adobe is advancing initiatives to protect content creators, including Content Credentials for attribution and "do not train" tags on shared images. The company also supports FAIR legislation to safeguard creators against impersonation.

Adobe recognizes that trust must be continuously earned and expressed gratitude for community feedback. It reaffirms its commitment to being a reliable partner for creators in the digital landscape.

As the new ToS is implemented, Adobe promises to engage with its user community to address concerns and protect rights. The company's efforts to rebuild trust and maintain its reputation as a responsible innovator are clear. Whether these moves will satisfactorily address user concerns remains to be seen, but Adobe appears committed to responding in good faith.