Unresolved outlier after early-March appeals — Antigravity / Gemini CLI / Gemini Code Assist still suspended

Google Antigravity
, ,
1

Hello,

I am posting this as one consolidated case because my situation now appears to be an unresolved outlier.

Summary:

  • Services affected: Antigravity, Gemini CLI, and Gemini Code Assist

  • First suspension and first appeal: both happened in early March 2026

  • Exact first date: I no longer have the exact day, but both happened in early March

  • Follow-up appeals: submitted multiple times afterward

  • Current status as of April 20, 2026: still suspended, no response, no reinstatement

What I already did:

  • Stopped using all third-party tools

  • Revoked all related third-party OAuth access

  • Cleared local tokens / sessions

  • Stopped all related automation

  • I am willing to use only official Google-supported access methods going forward

Additional context:

  • Google One support told me they do not have escalation access for ToS-related suspensions and that these cases are handled only through the appeal path

I understand the policy concern and I am not trying to argue about enforcement itself.

My question is narrower:
Could a Google team member please clarify whether this case is most likely:

  1. still pending review,

  2. missed by the automated reinstatement process,

  3. or classified as a second-violation / permanent-ban case?

If needed, I can provide masked account details and a concise appeal timeline privately.

@Abhijit_Pramanik @Mrinal_Ghosh
If this looks like a missed reinstatement or an unresolved outlier, could you please help confirm whether the correct team is tracking it?

2

Update :

Still suspended across Antigravity, Gemini CLI, and Gemini Code Assist.
No response to prior appeals, and no reinstatement yet.
All third-party access remains revoked.

If a moderator or Google team member needs masked account details, I can provide them privately.

3

Wow @Daniel_L,

I thought I was in a bad position. 17 days and counting.

Are you actually paying and getting no value?

For like 7 weeks already?

I struggle to understand how such behavior is good for business. Like why write that you’ll get a response after 1-3 days, and then make you wait more than 7 weeks?

What do they think will happen during those weeks? That you will not use other models? Not migrate to other IDE? Maybe even start paying to other providers instead of just waiting ..

Then what? Will developer come back once account gets unbanned? Or will they drift away and never come back?

I mean, the core logic is sound: you ban account, they confirmed that they will not authenticate third party coding agents, and in less than 24 hours you automatically reactivate account. If there is a second violation, then punishment should be more harsh, etc.

In my case, I had tested OpenCode already. Authenticated Copilot, Codex, etc. There is absolutely no intuition that it is not allowed, since I use of my quota regardless if I use Copilot CLI or OpenCode. And if it not allowed, then authentication should not be successful either.

And then I tested Pi Coding Agent, which allow me to authenticate with Antigravity. Next day I was banned. I had no idea it was against terms of agreement, so I appeal and confirm I will not do it again.

But to then remain banned for more than 17 days is beyond my reasoning ability. Especially when they ban paying customers, which obviously must result in loss of income when they do nothing to lift the ban.

And in your case, 7 weeks. I mean, wow, just woooow.

4

Yeah, that is pretty much the situation.

I was on Ultra when this first happened. After the account was not reinstated for a long time, I downgraded to Pro. This is still my primary Google account, but once the advanced AI-related benefits are removed or blocked, Ultra is simply no longer attractive enough for me to keep paying for it.

The reason I am still trying to find a channel to raise this, instead of just giving up and moving on, is that I still suspect this may be a system/process failure rather than an intentional long-term enforcement decision.

If Google clearly reviewed the case and decided this should be a permanent ban, that would be one thing. I would at least know where I stand. But what makes this frustrating is the silence: multiple appeals, no clear rejection, no reinstatement, and no way to know whether the case is pending, missed by an automated process, or actually classified as a permanent enforcement action.

That is why I keep framing it as an unresolved outlier. It feels like something that could be corrected if the right team actually looks at it. Of course, whether that is true is unknown. It is also unclear whether anyone at Google is still actively owning or tracking this specific issue anymore.

I agree with your point about the business side. If paying users lose access for weeks after acknowledging the policy concern and agreeing not to repeat the behavior, many of them will naturally migrate elsewhere. Even if the original enforcement logic makes sense, the lack of a reliable recovery path creates a much bigger trust problem.

5

Agreed.

And if they actually had a support team, someone whose responsibility is to ensure happy customers, then people like you and me would be able to reach an actual intelligent human being and explain that there is a flaw in their system.

My personal opinion is that they should first issue a warning, and then impose a ban, if the behavioral pattern continues. That would make far more sense, and allow most of the cases to be corrected before necessitating any dramatic and drastic actions.

The behavior we see here is very inhumane and surely done by AI developers, which rutinely “kill” agents that are “misbehaving”. But there is a fundamental difference between paying customers and “misbehaving agents”.

6

Are you still banned @Daniel_L?

Today, May 2nd, after 24 days, I noticed the ban has been lifted. After much suffering since April 8th.

Did I get any email telling me that I’ve been unbanned? No.

Any explanation why it took more than 72 hours? No.

In any case, my “Gemini Code Assist in Google One AI Pro” is now active again :heart:

7

happy to know that your account unbanned.

for me , still banned. just tried.