Enterprise AI Controls: Please Allow Model-by-Model Access

I work for a government agency that uses an Enterprise Creative Cloud plan for a workforce of over 1,000 employees.

With the recent rollout of third-party AI models across Adobe products, I noticed that attempting to use any model other than Firefly results in the following message:

“Your system administrator isn't allowing access to this feature right now” (see attached screenshot below)

After contacting our IT department, I learned that our agency has disabled access to all third-party AI models within Adobe. However, when I asked whether access could be enabled for a specific, pre-approved model (in this case, Google Gemini / Nano Banana), I was told this was not possible.

According to our IT team, Adobe’s current enterprise controls only allow all-or-nothing access to third-party AI models — meaning administrators must either enable access to all external models or disable them entirely.

This limitation makes it difficult for large organizations with strict security policies to adopt Adobe’s new AI integrations.

Feature Request:
Please allow Enterprise Creative Cloud administrators to manage third-party AI access with more granular controls, including:

• Enabling or disabling specific AI models individually (e.g., allow Gemini but not others)
• Granting access on a user-by-user or group basis

This level of administrative control would make it significantly easier for enterprise and government organizations to safely adopt Adobe’s expanding AI ecosystem.