Anthropic has temporarily banned the creator of OpenClaw from accessing its Claude AI models, highlighting growing tensions between AI platform providers and third-party developer tools. The move comes as the company tightens enforcement around how its models can be accessed and used, particularly through unofficial integrations and agent-based systems.
The restriction is part of a broader policy shift by Anthropic to block third-party tools from using Claude through subscription-based access. The company has implemented technical safeguards to prevent unauthorized integrations, especially those leveraging OAuth tokens to connect external tools like OpenClaw. These measures are aimed at ensuring compliance with its usage policies and maintaining control over how its AI systems are deployed.
OpenClaw, an open-source AI agent platform, has gained popularity for enabling advanced automation workflows by connecting large language models to external applications and services. However, such tools can significantly increase compute usage compared to standard chatbot interactions, placing additional strain on infrastructure. This has been cited as one of the key reasons behind Anthropic’s stricter stance on third-party access.
The situation has also been influenced by recent security-related developments, including the leak of Anthropic’s Claude Code source components, which raised concerns about vulnerabilities and misuse. In response, the company has taken a more controlled approach to its ecosystem, prioritizing security and stability over open integrations.
This development reflects a wider industry trend where AI companies are increasingly limiting open access to their models while shifting toward more controlled and monetized usage frameworks. As AI tools become more powerful and resource-intensive, companies like Anthropic are balancing developer flexibility with the need to manage risks, infrastructure demands, and long-term sustainability of their platforms.
