Back to List
Industry NewsAICloud ServicesDeveloper Tools

Google Restricts Antigravity Access for OpenClaw Users Citing 'Malicious Usage' and Overwhelmed Systems, Highlighting Rivalry with OpenAI

Google has sparked controversy by restricting access to its Antigravity 'vibe coding' platform for users, particularly those integrating with the open-source AI agent OpenClaw. Google alleges 'malicious usage,' stating that these users were accessing an excessive number of Gemini tokens through third-party platforms like OpenClaw, leading to service degradation for other Antigravity customers. Some affected users reported losing access to their Google accounts. This move is seen as a strategic response, especially given that OpenClaw's creator, Peter Steinberger, recently joined OpenAI, Google's primary rival. While OpenClaw remains open-source, it is now financially backed and strategically guided by OpenAI. Google DeepMind engineer Varun Mohan confirmed the crackdown, noting the need to address service degradation caused by users not adhering to the Terms of Service, and indicated a path for some unaware users to regain access.

VentureBeat

Google has initiated a significant enforcement action against certain users of its Antigravity 'vibe coding' platform, citing 'malicious usage' and causing considerable controversy among developers. The restrictions, which began this weekend and continued into Monday, February 23rd, primarily affected users who had integrated the open-source autonomous AI agent OpenClaw with Antigravity-built agents, or those who had connected OpenClaw agents to their Gmail accounts. These users subsequently reported losing access to their Google accounts.

According to Google, the affected users were leveraging Antigravity to obtain a larger volume of Gemini tokens via third-party platforms such as OpenClaw. This activity, Google claims, overwhelmed its system and degraded the quality of service for other Antigravity customers. The company's action has effectively cut off several users, bringing to light potential architectural and trust issues associated with OpenClaw's integration with Google's services.

The timing of Google's crackdown is particularly noteworthy. Just a week prior, on February 15th, OpenAI CEO Sam Altman announced that Peter Steinberger, the creator of OpenClaw, had joined OpenAI to lead its 'next generation of personal agents.' Although OpenClaw continues to operate as an open-source project under an independent foundation, it now receives financial backing and strategic guidance from OpenAI, Google's main competitor in the AI space. By severing OpenClaw's access to Antigravity, Google is not merely safeguarding its server infrastructure; it is also effectively disrupting a channel that allowed an OpenAI-affiliated tool to utilize Google's advanced Gemini models.

Varun Mohan, a Google DeepMind engineer and former CEO and founder of Windsurf, addressed the situation in an X post. He stated that the company had observed a 'massive increase in malicious usage' of the Antigravity backend, which had severely impacted the quality of service for legitimate users. Mohan emphasized the necessity of quickly restricting access for users who were not using the product as intended. He also acknowledged that a subset of these users might have been unaware that their actions violated Google's Terms of Service (ToS) and indicated that a pathway would be provided for them to regain access.

Related News

Meituan Showcases AI Innovations at ACL 2026: Advancing LLM Evaluation, Reasoning, and Generative Recommendations
Industry News

Meituan Showcases AI Innovations at ACL 2026: Advancing LLM Evaluation, Reasoning, and Generative Recommendations

The Meituan technical team has achieved significant recognition at the ACL 2026 conference, with six papers accepted into this premier international forum for computational linguistics and natural language processing. These research contributions span critical frontiers in the AI landscape, including large language model (LLM) capability evaluation, complex process reasoning, and the optimization of competition-level mathematical thinking. Additionally, the papers explore advancements in reinforcement learning and the evolution of generative recommendation systems. By addressing these diverse technical directions, Meituan is actively shaping a new paradigm for generative AI, focusing on bridging the gap between theoretical research and practical industrial applications. This selection of papers highlights Meituan's commitment to enhancing model intelligence and reasoning capabilities to solve sophisticated real-world problems.

Meituan LongCat Releases General 365: A New Benchmark for AI Reasoning Evaluation
Industry News

Meituan LongCat Releases General 365: A New Benchmark for AI Reasoning Evaluation

Meituan's LongCat team has officially launched General 365, a rigorous new benchmark designed to evaluate the reasoning capabilities of large language models. In a comprehensive test of 26 mainstream models, the results revealed a significant performance gap in the industry. Even the top-performing model, Gemini 3 Pro, achieved an accuracy rate of only 62.8%. Furthermore, the vast majority of the models tested failed to reach the 60% threshold, which is considered the passing mark for this evaluation. This release sets a challenging new standard for AI development, highlighting that complex reasoning remains a major hurdle for even the most advanced artificial intelligence systems currently available.

Managing AI-Driven Development: Meituan’s Strategy for Refactoring 310,000 Lines of Code Using Agent Evaluation Logic
Industry News

Managing AI-Driven Development: Meituan’s Strategy for Refactoring 310,000 Lines of Code Using Agent Evaluation Logic

Meituan's technical team has shared a comprehensive analysis of their experience refactoring 310,000 lines of code in an environment where over 90% of code is AI-generated. The core insight is that while AI significantly accelerates code production, it can also amplify technical debt and systemic chaos without proper constraints. To mitigate this, the team adopted an 'Agent evaluation' mindset to manage AI coding. By implementing a framework consisting of technical debt sorting, rule construction, standardized operating procedures (SOPs), and a Pre-PR (Pull Request) mechanism, they successfully transformed large-scale refactoring from a high-cost, specialized effort into a continuous, daily iterative process. This approach ensures that AI remains a productive tool rather than a source of unmanaged complexity.