As @tuapuikia99, you probably aren’t.
i think the Opencode guy on twitter did tease that they are working with google to get access in their stuff , https://x.com/thdxr/status/1992011763475476642
Curious
It’s truly curious and reinforces the shortcomings in the duty of information and transparency that we discussed. The image shows that the link to the Gemini CLI’s “Terms of Service and Privacy Notice” on GitHub is broken (404 - page not found).
- Lack of Clear Rules
The tool’s installer directs the user to a document that simply doesn’t exist or has been removed.
How can a user be banned for violating terms that Google itself (or the maintainer of the official repository) doesn’t make available for reading at the time of installation?
- Contradiction with OAuth Policies
Google’s official OAuth policy requires applications to provide valid links to their Terms of Service and Privacy Policy on the homepage.
A broken link (404) in an official or widely used repository violates this security and transparency guideline of Google itself.
“Beyond the technical contradictions, the official link to the Gemini CLI Terms of Service results in a 404 error. Google is banning paying users based on rules that it itself fails to make accessible and transparent.”
It is true that we reverse-engineered the OAuth process, but your marketing materials clearly stated: “Purchase Google One and start using.”
If you provide a certain quota and then flag users for “abuse” when they actually utilize it, where exactly does the responsibility lie? If high-volume usage causes operational losses, the correct approach should be to upgrade your quota tracking system—similar to how burstable performance VPS providers manage resources—and update your Terms of Service (ToS) while proactively notifying every user about when and how the service changes.
Furthermore, you should provide a clear process for handling those who do not agree to new terms. For example, for users who have already committed to a 1-year subscription, you should either honor the original terms until the subscription expires or offer a pro-rated refund for the unused portion of the service.
Even in cases of “abuse” bans, an email notification should be sent at the very least. Punishment without prior warning is fundamentally unfair.
I ended up discovering a bug that may be influencing this, related to Google Cloud Workspace and Google Personal Account.
I just wanted my antigravity to work.
*Update: Evidence of False Positive ToS Bans tied to GCP Identity/Session Desync (Even with Native CLI)*
Hi everyone, just wanted to share a critical update on my investigation into the sudden 403 ToS bans impacting Paid/Ultra Tier users.
After migrating completely away from any third-party open-source CLIs and running **exclusively on the native official CLI**, the severe account instability and 403 ToS blocks persisted.
However, while doing a deep audit of my Google Cloud environment, I stumbled upon a major architectural anomaly that I believe is the actual root cause triggering the automated WAF/Anti-Abuse system.
It appears there is a severe “Session Bleed” or identity desynchronization occurring in the backend when a user operates a standard Consumer Account (e.g., standard Google account) alongside a Workspace/Cloud Identity organizational account in the same environment.
When the official native CLI attempts the OAuth handshake, the Application Default Credentials (ADC) or the active session token gets polluted with an inconsistent mix of organizational constraints and consumer tier limits.
**The Result:** The automated security backend reads this “schizophrenic” token state not as a legitimate Google Cloud UI/CLI bug, but erroneously flags it as a “Violative/Unauthorized Client” or an attempt to bypass ToS, triggering an immediate and silent 403 ban on the Gemini API.
This means many of us aren’t being banned for “abusing” the API or using “unauthorized tools.” We are victims of a False Positive generated by a session management bug within GCP’s own identity infrastructure crossing wires between our Consumer and Workspace profiles.
I am currently compiling a formal architecture vulnerability report for the Google Cloud VRP (Vulnerability Reward Program) regarding the underlying GCP Identity flaw. But the AI team needs to urgently halt these automated ToS bans on Paid Tier accounts until the Cloud Infrastructure team resolves this cross-account authorization bleed.
The collateral damage on legitimate developers is entirely unjustified.
-–
*(Note: I am holding back the exact reproduction steps of the GCP bug as it is being routed to the Cloud VRP, but this context is vital for the AI team to understand the false positives).*
It’s stuff like this, and all the free EDU accounts, which are really hurting Antigravity.
Don’t get me wrong, I agree, I want a service that lets me do what I want with the tools I want..
… but that’s not Antigravity… That’s Google AI Studio and/or VertexAI.
Yes, I want the same functionality, but I didn’t sign up until after I accepted this fact…
…and it’s the same with Claude Code.
—
If you want to not be locked into a specific client, you have to stop supporting those clients, or the companies which use this type of service pattern. Instead, find providers which align with your objective and beliefs.
You will quickly see that Claude Code and Antigravity both have (at least with what is available today) increased quota, but they do that inside a “walled garden”. If you want more features, you gotta upgrade.
Do you have to continue paying/supporting? No.
—
Furthermore, you should provide a clear process for handling those who do not agree to new terms. For example, for users who have already committed to a 1-year subscription, you should either honor the original terms until the subscription expires or offer a pro-rated refund for the unused portion of the service.
This is absolutely true. It also includes how Claude went from 5-hour to 3-day/5-day/whatever MID-MONTH/PAYMENT.
So, I say the same thing to Google that I just said to you, you can’t just do what you want, when you want, especially when people are paying…
- You can’t neuter quota and accessibility mid-subscription
- You can’t use someone’s service in ways they don’t want you to use it
all the event just like Fruit of the Poisonous tree.
I still don’t understand why the Official Oauth test break the rule.
If I am a developer, It provide the Official Way to use the OAuth to do some simple test,
And then got banned forever? Quite Strange and unfair.
If the way to connect the OAuth is not allowed, you should disable it.
Still feel sick about this rough decision.
Hey Google Developers Team, I tried the appeal via the email, but did not get a response. I became aware of this thread today, so i’m writing here to appeal the ban applied to my account.
Account:
I want to be fully transparent about what happened, as I believe this was an honest and unintentional mistake on my part.
I was using OpenCode, an AI coding tool, alongside a community plugin called “Oh My OpenCode”. Since I was not authenticated via Google Auth on the OpenCode CLI directly, the Oh My OpenCode plugin prompted me to log in — and I did so through that prompt, without realizing this was not an officially supported or authorized authentication flow.
The key context here is: OpenCode supports OAuth for several other providers, so when the plugin presented a Google login prompt, I genuinely assumed it was a permitted integration. I had no awareness that Google Auth was not sanctioned for use with OpenCode, and I had no intent to circumvent any terms of service or abuse Google’s authentication infrastructure in any way. I was simply following a login prompt presented to me by a tool I trusted.
I take full responsibility for not verifying whether that specific flow was compliant before proceeding. That was a mistake, and I deeply regret it. Had I known it violated Google’s policies, I would never have used it. I have since stopped using the plugin and will ensure any future authentication flows are explicitly verified against Google’s Terms of Service before use.
Being banned from Gemini CLI and Google AI Studio has significantly impacted my development workflow, as these are tools I rely on daily for legitimate personal and professional projects.
I humbly ask that you consider:
-
The unintentional nature of the violation
-
That the login was initiated by a third-party plugin prompt, not a deliberate attempt to misuse Google Auth
I would be truly grateful for a second chance, the ban caused a massive hurt in my development productivity. Please reinstate my account access at your earliest convenience.
Thank you so much for your time and consideration. I’m happy to provide any additional information or clarification that may help with this review.
Warm regards,
Tulio Pinheiro
I fully support the ban. You clearly didn’t read the terms of service before using the product.
I didn’t even want to reply to this AI-generated slop, but I have to point out just how obvious and clueless you are. Google AI Ultra gives you high limits for specific services: Gemini Code Assist, the official Gemini CLI, and Antigravity. Those services are meant to be used strictly on their specific, official applications, not through reverse-engineered apps you randomly picked up on GitHub, like OpenClaw.
You are directly breaching Section 6 of the terms: ‘You must not abuse, harm, interfere with, or disrupt the Service’. The terms explicitly forbid using the service in connection with products not provided by Google. You are downloading and connecting unauthorized software without even understanding what it does to your authentication tokens.
You aren’t being banned for ‘using open-source tools’; you are being banned for utilizing third-party software that explicitly violates the ToS. Your massive legal argument about Brazilian consumer law falls completely flat when you didn’t even read what you agreed to in the first place.
It is genuinely funny to read the ‘Discrimination against Open Source’ complaint, especially since Antigravity is a proprietary service.
I love that the AI-style responses in this thread are just trying to validate you instead of telling you the harsh truth: you are completely clueless, you broke the rules, and you deserve the ban.
Google punished everyone without investigation. I never used OpenClaw (or similar things) but also received the limitations
You clearly didn’t read the entire post.
I read the entire post, Rodrigo. I read your AI-generated , and honestly, I doubt you even read it yourself before hitting reply. It’s hilariously wrong. Generating a massive essay about the Brazilian General Data Protection Law doesn’t magically erase the fact that you explicitly confessed to using OpenClaw in your very first paragraph. You broke Section 6 of the terms. The length of your excuses doesn’t change the facts.
1. Here, consumers are expected to actually read the contract before throwing $250 a month at a service.
2. You don’t get ‘technical support’ when you get caught explicitly violating Section 6 of the Antigravity Additional Terms of Service.
3. Your login isn’t in an ‘inconsistent state’ because of some magical server bug. Your account is permanently blacklisted by the security backend because you abused the system with unauthorized third-party tools.
Google engineers don’t need a ‘forensic examination’ of your computer, Rodrigo. I did my own ‘forensic examination’ of your posts, and I’m providing the results to you for free right here.
The authentication process isn’t ‘erratic’ because of Firefox or your Workspace account; it is permanently blocking you because you violated the exact terms highlighted in this image.
I don’t need to work for Google to understand something this incredibly obvious. The difference between us isn’t technical skill, it’s that I can actually read a binding contract and use basic common sense before blindly connecting a $250/month AI ultra account to reverse-engineered GitHub scripts. The evidence of your violation isn’t on your hard drive, it’s in your own forum posts. Give it a rest.
How do I explain what, Rodrigo? Your inability to take a readable screenshot? Or why you are desperately trying to blame Google Cloud for a ToS ban on a consumer AI subscription?
Hello everyone,
I’m in the same situation — my account was disabled around Feb 12 with the “This service has been disabled in this account for violation of Terms of Service” message.
Like many here, I had used third-party CLI tools (OpenClaw-based) without fully realizing this was against the ToS. I understand the concern from Google’s side and have since taken corrective steps:
- Deleted all Gemini API keys from Google AI Studio
- Revoked all Gemini-related OAuth connections (Gemini, Gemini CLI Workspace Extension, Gemini Code Assist and Gemini CLI, Google AI Studio, Google Antigravity)
- Verified all my Google Cloud Console projects are clean
- Confirmed no third-party tools retain any access
I emailed gemini-code-assist-user-feedback@google.com on Feb 13 and sent a follow-up today (Feb 26), but haven’t received a response yet.
I’d really appreciate any guidance on the reinstatement process or expected timeline. Thank you for your time and attention to this.
