
Anthropic officially bans using subscription authentication for third-party Claude use
Anthropic has enacted a major policy change, now prohibiting the use of subscription-based OAuth authentication for any third-party products, tools, or services. This update specifically bans OAuth tokens from Claude Free, Pro, and Max subscriptions in all contexts outside of Claude Code and Claude.ai platforms.
These restrictions were detailed through an updated “Legal and compliance” resource doc, clarifying that developers integrating Claude’s capabilities, including those using the Agent SDK, must now use API key authentication issued through Claude Console or a supported cloud provider. Following this revision, third-party developers are no longer allowed to offer Claude.ai login or to route requests on behalf of users using Free, Pro, or Max plan credentials.
This development immediately impacts teams and toolmakers who previously bypassed usage-based API fees by employing Claude Max subscriptions with integrated third-party solutions. For example, OpenClaw and OpenCode had provided users the ability to rely on personal subscriptions instead of per-usage charges. While this practice had existed in a gray area, the documentation update makes clear such patterns are no longer permitted.

Comments
So, we're now at phase 2 of 3 of enshittification: "they abuse their users to make things better for their business customers".
Theses days, AI enshittification evolves faster than AI.