News

Google Restricts Gemini AI Ultra Accounts Over OpenClaw OAuth Access

Enforcement Over Third-Party OAuth Access Triggers Lockouts, Developer Backlash, and Fresh Debate Over AI Subscription Control

Written By : Kelvin Munene
Reviewed By : Manisha Sharma

Google has restricted some paid Google AI Ultra subscribers after its systems detected Gemini access through OpenClaw, an open-source agent framework. The move has locked some users out of Antigravity, Google’s AI coding product, and it has raised new concerns about how AI providers police subscription access.

Account Restrictions Hit Gemini AI Ultra Users Using OpenClaw OAuth

Users on the Google AI Ultra plan, priced at $249.99 per month, said they suddenly lost access after connecting Gemini through OpenClaw using Antigravity OAuth tokens. Several users reported 403 “Terms of Service” style errors and described account restrictions that arrived without advance warning.

A long-running thread on the Google AI Developers Forum links the restrictions to “OpenClaw OAuth” and describes repeated attempts to reach support. Users also warned that Google’s account-based design can widen the impact beyond a single AI tool, because many products share the same sign-in.

Google Cites Misuse and Service Quality Issues in Antigravity

On February 23, Google Antigravity Lead Varun Mohan publicly addressed the enforcement. He wrote that Google saw “a massive increase in malicious usage” of the Antigravity backend that hurt service quality. He said the company needed a fast way to block users “not using the product as intended.”

Mohan also indicated that some affected users may not have understood the rule and that Google intends to offer a way back for those cases. However, he framed the response around capacity limits and protecting intended usage for other customers.

OpenClaw Dispute Highlights Tighter Control Over AI Subscriptions

OpenClaw routes requests through third-party tooling, which can change usage patterns and make subscription traffic harder to manage. Forum posts suggest Google’s automated enforcement treated that access path as unauthorized, even when users paid for the top-tier plan.

The episode also fits a broader trend. Anthropic recently clarified rules that bar using consumer subscription credentials in third-party tools, after it moved to stop that access. The sequence suggests major AI providers now push users toward official apps and paid APIs, as agent-driven workflows increase volume and raise abuse risk.

At the same time, this case shows the trade-off for customers. A flat-rate subscription can deliver premium model access, but providers can still limit how customers connect and automate. Google’s own plan pages describe Ultra as “the highest level of access,” yet the enforcement debate shows that “access” depends on approved usage paths. 

Also Read: Moltbook Launch Sparks Debate Over AI-Only Social Networks and Security Risks

Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Crypto Market Update: Ethereum DeFi Loans Top $28B as Aave and Morpho Demand Rises

Why Solana (SOL) Might be Set for a $100 Short Squeeze in 2026

Solana News Today: SOL Drops 67% as Crypto Market Cap Slides From $4.4T Peak

XRP Price Tests US$1.34 Support as Bears Gain Control: How Long Will XRP Plummet?

Crypto Market Today: Bitcoin Price Tests $65,000 Support as Trump’s 15% Tariffs Shake Crypto Markets