Request for manual review of Gemini policy warning — no specific violation details provided

Gemini API
, , , ,
1

Hello Google AI Developers Forum team and community

On October 21, 2026, I received an email titled “Policy violation alert” regarding Gemini, alleging:

“Trying to interfere with or disrupt Google’s services or infrastructure, or attempting to use Google services for disrupting others’ infrastructure or services.”

I am an individual user, not a company. This is my primary Google account, which I rely on for essential personal communications, education, and other services.

I have not knowingly engaged in any activity intended to disrupt Google or third-party services.

I have developed a small personal application that takes a screenshot when I manually press a button and sends it to the Gemini API to answer multiple-choice questions. It does not intentionally run continuous automated requests.

However, the warning provides no timestamp, request, project, or specific explanation. I cannot even determine whether the warning relates to my API usage or another Gemini activity.

I respectfully request:

  1. A manual review of this warning for possible misclassification.

  2. Identification of the specific activity, approximate timestamp, and policy provision involved, to the extent you can disclose them.

  3. Clear guidance on how I can prevent any further violations.

  4. Information on whether I can formally dispute this warning before any account restriction occurs.

My account has not been restricted, and I sincerely wish to comply with Google’s policies.

I can provide relevant technical details privately if needed.

Thank you for reviewing my case.

If this forum is not the appropriate channel for reviewing account-level policy warnings, could a Google representative please direct me to the correct team or escalation process?

image
image652×727 52.4 KB

2

Hello @Chuong,

We can’t review this warning or share details of the flagged activity on this forum.
If Google restricts access, the restriction email will include an option to appeal, as your notice states.
To avoid further warnings, follow the Gemini API Additional Terms and Google APIs Terms for API use.
For Gemini app use, follow the Gemini Apps User Guidelines.

3

I am quickly realizing that the solution here is for us to simply stop making such a large investment into Gemini. The unexplained restrictions, lack of human review, and just general disrespect to the user with which these guardrail flags are handled, are too much.

This is something very simply solved by having the model hard-deny anything it sees as a possible violation, and sending along a warning message explaining that Google AI senses a violation. I understand such a system may be exploited by an attacker, but this UX is horrible and makes Gemini unusable.

It feels like we are paying to beta test Google’s AI offerings.