Anthropic has officially pulled the plug on third-party AI agent access to the Claude subscription, marking a significant shift in how users can leverage its models outside the company’s native ecosystem.
According to Anthropic Claude Code exec Boris Cherny, starting today, April 4, at 12 p.m. PT (8 p.m. BST), Claude Pro and Max subscribers can no longer use their existing subscription limits to power third-party automation tools like OpenClaw, a popular open-source AI agent framework widely used for tasks including email management, web browsing, and smart home automation. Enforcement begins with OpenClaw and will be rolled out to all third-party harnesses in the coming weeks.
Anthropic’s Consumer Terms of Service have technically prohibited unauthorized third-party tool access since at least February 2024, but enforcement was lax for years.
Many OpenClaw users exploited an OAuth authentication loophole, the same login method used by Claude Code, to pipe subscription-tier Claude models into their personal agents at a flat monthly rate.
Anthropic formally revised its terms in February 2026 to close this gap, explicitly stating that OAuth authentication is now reserved exclusively for Claude Code and Claude.ai.
The crackdown was triggered, in part, by the scale of the problem. Anthropic stated in its notification email that third-party harnesses were placing an “outsized strain” on its infrastructure and that capacity must be prioritized for customers using its core products.
Steinberger says that he and OpenClaw board member Dave Morin “tried to talk sense into Anthropic, best we managed was delaying this for a week.”
Users who wish to continue integrating third-party agents with their Claude account now have two options: enable pay-as-you-go “extra usage” billing that is charged separately from their subscription, or authenticate using a standard Claude API key with metered API pricing.digitaltrends
To soften the financial blow, Anthropic is offering a one-time credit equal to the user’s monthly subscription cost, redeemable by April 17, along with discounts of up to 30% when pre-purchasing bundles of extra usage. Subscribers who prefer not to continue under the new terms will also receive a full subscription refund option.
The move has provoked significant pushback from the developer and power-user community. Many users who relied on OpenClaw-plus-Claude workflows report per-interaction costs now ranging from $0.50 to $2.00 per agent task, making autonomous agent use cases economically unviable for hobbyists and solo developers.
Critics argue that Anthropic marketed agentic workflows while simultaneously restricting the most affordable path to build them.
The policy shift underscores a broader industry tension between AI companies monetizing infrastructure and developer communities seeking open, flat-rate access to frontier models.
Follow us on Google News, LinkedIn, and X for daily cybersecurity updates. Contact us to feature your stories.
