Google has recently suspended numerous users of the open-source tool OpenClaw from accessing its Antigravity AI platform. This decision has ignited controversy due to Google’s strict enforcement of its terms of service.
Reasons for the Suspension
The suspension primarily affects developers who utilized OpenClaw’s OAuth plugin to exploit subsidized tokens of the Gemini model, leading to backend overloads and service disruptions. OpenClaw, which gained significant traction since its launch in November 2025, offered local AI agents for tasks such as email management and web browsing. Users bypassed official channels to access premium models like Gemini 2.5 Pro at lower costs, breaching Google’s terms of service.
Google identified these activities as violations, as they utilized Antigravity’s infrastructure for non-Antigravity products, triggering automated flags for “malicious usage” patterns.
Impact on Developers
By mid-February 2026, reports emerged on Google’s AI Developer Forum about AI Ultra subscribers, who pay $249.99 monthly, encountering sudden access denials and full account restrictions without prior warning. The impact extended beyond Antigravity and Gemini CLI access, affecting services such as Gmail and Workspace, causing significant distress among users.
Varun Mohan, who leads Google’s DeepMind product initiatives, addressed the matter publicly, acknowledging the service quality decline due to such integrations. He assured a limited reinstatement path for those unaware of their violations but maintained the necessity to prioritize legitimate users.
Industry Reactions and Future Outlook
This incident mirrors Anthropic’s recent policy update, which forbids third-party OAuth integration in tools like OpenClaw due to token misuse and debugging challenges from unexpected traffic spikes. Cybersecurity specialists have highlighted OpenClaw’s vulnerabilities, noting over 21,000 instances susceptible to security threats such as infostealers and supply chain attacks.
China’s industry ministry has also expressed concerns about risks from misconfigured systems, emphasizing the potential for data breaches. OpenClaw’s creator, Peter Steinberger, criticized Google’s actions as excessive and announced plans to cease Antigravity support, highlighting a shift towards alternatives like Nanobot and IronClaw.
As the competitive landscape evolves, AI agents’ growing presence necessitates careful consideration of terms of service and security implications. Developers are increasingly challenged to balance convenience with compliance, as unauthorized access can lead to significant disruptions.
For ongoing updates on cybersecurity and AI developments, follow us on Google News, LinkedIn, and X. Reach out to feature your stories with us.
