사전 경고 없는 계정 제한 — Google AI Ultra / OpenClaw OAuth  | GeekNews
News

사전 경고 없는 계정 제한 — Google AI Ultra / OpenClaw OAuth | GeekNews

fanotify
2026.02.24
·News·by 성산/부산/잡부
#Account Suspension#Customer Support#Google AI#OAuth#Terms of Service

Key Points

  • 1Google AI Ultra/Pro paid subscribers are facing unilateral account suspensions without warning after integrating third-party tools like OpenClaw via OAuth, which Google deems a Terms of Service violation for misusing its Antigravity servers.
  • 2Google enforces a "zero tolerance policy," refusing account recovery for these violations, while users report systemic customer support failures, including unresponsiveness, department transfers, and forum account bans, often with continued billing for suspended services.
  • 3This mass ban has led to widespread user frustration, data migration efforts, and a strong push towards competitor services, with many criticizing Google for its disproportionate punishment, lack of transparency, and perceived betrayal of its paying customer base.

A mass account suspension event affected Google AI Ultra and Pro paid subscribers who integrated their accounts with OpenClaw via OAuth. Google's internal investigation concluded that this integration constituted a violation of its Terms of Service (ToS). Specifically, Google identified the use of authentication information via OpenClaw as leveraging Antigravity servers for non-Antigravity products. This was classified as a "zero tolerance policy" violation, leading to irreversible account suspensions without prior warning or communication.

The core methodology cited for the ban revolves around the unauthorized use of Google's internal API infrastructure, specifically Antigravity, which is designed for first-party Google applications (e.g., Gemini CLI, Cloud Code Private API). When users connected their Google AI accounts to OpenClaw using OAuth, it appears OpenClaw extracted and reused internal OAuth tokens. This allowed the third-party tool to access Google's backend services, likely bypassing the intended consumption model for Google AI services. The concern stems from the difference in how first-party (optimized, potentially with caching) versus third-party (less optimized, potentially higher resource consumption) tools interact with the backend. It implies that OpenClaw's method might have enabled effectively "unlimited calls" that would otherwise incur significant metered API costs, thus circumventing Google's economic model for its services. The detection mechanism is suggested to be pattern-based, identifying abnormal usage patterns indicative of a ToS violation.

Users impacted by the ban faced a complete loss of access to their Google AI Ultra/Pro subscriptions, Antigravity services, Gemini CLI, and Cloud Code Private API, while continuing to be billed. Customer support channels were largely ineffective, characterized by department redirects, unresponsiveness, and a lack of official communication. Critically, Google employees reportedly deleted forum posts discussing the issue and banned forum accounts of users attempting to inquire about the problem, indicating a deliberate suppression of information. Affected users expressed strong dissatisfaction with Google's unilateral actions, perceived betrayal of paid customers, and a complete breakdown of support, leading many to migrate away from Google's ecosystem and consider legal action. The incident highlights concerns about platform control, data access rights, and the potential for a large tech company's arbitrary enforcement policies to impact users' entire digital lives.