Anthropic shuts down free claude access, sparks investor concerns
Anthropic, already navigating choppy waters after a Pentagon dispute, has abruptly ended free access to its AI assistant, Claude, for third-party applications like OpenClaw. The move, announced via a flurry of posts on X by Boris Cherny, the executive overseeing Claude Code, effectively forces users to pay a subscription fee to continue utilizing the model – a development that’s raised eyebrows and triggered fresh questions about the company’s strategy.
A sudden shift in claude's accessibility
Starting April 4th, Anthropic’s subscriptions will no longer cover integration with third-party tools. Cherny cited the company’s struggle to meet surging demand for Claude as justification. “Our subscriptions weren’t designed for these third-party tool usage patterns,” he stated, suggesting a misalignment between the model’s intended use and the way developers were leveraging it. Users wishing to continue accessing Claude through platforms like OpenClaw now face the prospect of an additional subscription, albeit with a discounted rate.
The timing is particularly noteworthy. Just weeks prior, Peter Steinberger, the creator of OpenClaw, opted to pivot to OpenAI, transforming OpenClaw into an open-source project backed by Sam Altman’s company – a direct competitor to Anthropic. Steinberger, also on X, detailed attempts to negotiate with Anthropic, claiming they “managed to get a one-week delay at best.” He added a pointed observation: “It’s curious how things line up: first they copy some popular features into their closed system, then they block access to the open-source code.” The sequence of events paints a picture of a company tightening its grip on its Technology, potentially stifling innovation in the open-source community.
The implications extend beyond mere accessibility. While Anthropic has attempted to deflect criticism, framing itself as a “big supporter of open source” and even mentioning contributions to OpenClaw’s cache efficiency, the move has fueled speculation about its long-term vision. Is this a calculated maneuver to monetize Claude more aggressively, or a sign of deeper strategic anxieties?
Despite the backlash, Claude’s team has offered some responses. Cherny emphasized Anthropic’s commitment to open source, highlighting his own contributions to OpenClaw’s efficiency. But the gesture feels hollow against the backdrop of the imposed restrictions. The swiftness and severity of the change suggest that, for all the talk of open collaboration, Anthropic is prioritizing control and profitability, even if it means alienating some of its earlier adopters. The AI bubble, it seems, has found a hard limit in the realities of human economics and corporate ambition.