OpenAI vs. Claude: Inside the Latest API Access Dispute Fueling AI Industry's Fiercest Rivalry

Despite headlines describing an "OpenAI vs. Claude legal battle," there is no active lawsuit pitting Anthropic, the maker of Claude, directly against OpenAI in court. The most significant clash between the two companies remains a terms-of-service dispute from last year, when Anthropic cut off OpenAI's access to Claude's API, an episode that continues to shape how the rival AI labs discuss access, competition and industry norms even as each company separately fights its own legal battles over copyright and government contracting.

Anthropic revoked OpenAI's API access to its models in a decision made public in early August 2025, according to multiple sources who spoke with WIRED at the time. Anthropic said the move came after it discovered OpenAI's technical staff had been using Claude's coding tools ahead of the launch of GPT-5. "Claude Code has become the go-to choice for coders everywhere, and so it was no surprise to learn OpenAI's own technical staff were also using our coding tools ahead of the launch of GPT-5," Anthropic spokesperson Christopher Nulty said in a statement to WIRED. "Unfortunately, this is a direct violation of our terms of service."

According to Anthropic's commercial terms of service, customers are barred from using the company's models to "build a competing product or service, including to train competing AI models" or to "reverse engineer or duplicate" its services. Sources told WIRED that OpenAI had been integrating Claude into its internal development systems through special API access rather than the standard chat interface, allowing the company to evaluate Claude's performance in areas including coding, creative writing and safety-related scenarios, including sensitive categories such as CSAM, self-harm and defamation, and to use that data to inform adjustments to its own models.

OpenAI pushed back on the characterization of its conduct as improper, framing the practice as standard across the AI industry. "It's industry standard to evaluate other AI systems to benchmark progress and improve safety," OpenAI chief communications officer Hannah Wong said in a statement. "While we respect Anthropic's decision to cut off our API access, it's disappointing considering our API remains available to them." Anthropic said it would continue to ensure OpenAI retained some API access specifically for benchmarking and safety evaluation purposes, consistent with what it described as standard industry practice, though it did not detail how that access would function under the broader restriction.

The dispute was not without precedent for Anthropic. The company had cut off API access for Windsurf, a coding startup, in June 2025, without publicly specifying a reason, though the move was widely believed to be linked to rumors that Windsurf was a potential acquisition target for OpenAI. That deal ultimately fell through.

Since that August 2025 episode, tensions between major AI labs over access and competitive positioning have continued to surface in other forms, though none have escalated into direct litigation between OpenAI and Anthropic specifically. More recently, Anthropic has found itself in a separate access dispute with a different company entirely. Chinese e-commerce giant Alibaba announced it would bar employees from using Anthropic's AI tools for work purposes starting July 10, 2026, placing Claude Code on a high-risk software list over concerns about potential security vulnerabilities, according to CNBC. That move followed a June 2026 letter Anthropic sent to the U.S. Senate Committee on Banking, Housing, and Urban Affairs accusing Alibaba of "brazenly" and "illicitly" attempting to extract its AI capabilities, part of Anthropic's broader effort to close loopholes that had allowed companies in restricted countries to access Claude through third-country intermediaries.

Where actual courtroom litigation involving Anthropic does exist, it centers overwhelmingly on copyright rather than any direct conflict with OpenAI. In Bartz et al. v. Anthropic, a class of authors sued the company alleging it illegally used their books, including copies downloaded from pirate websites, to train Claude. A federal judge ruled that training AI models on copyrighted books can constitute fair use, but that storing pirated copies obtained through so-called shadow libraries does not. Following that ruling, Anthropic agreed to a $1.5 billion settlement, with an estimated payout of roughly $3,000 per affected work, a deal that received final approval following a fairness hearing in May 2026.

OpenAI faces its own extensive slate of copyright litigation, largely separate from Anthropic's cases, including a closely watched lawsuit brought by The New York Times and consolidated class-action claims from a group of authors and organizations including The Authors Guild, all currently proceeding through multidistrict litigation in the Southern District of New York. Legal analysts have suggested that Anthropic's settlement could create pressure on OpenAI to reach its own resolution in that litigation, given the precedent it establishes for how courts and plaintiffs are approaching AI training data disputes industrywide.

Separately, Anthropic has been engaged in a distinct legal fight with the U.S. government rather than with any private competitor. The company sued Defense Secretary Pete Hegseth and the Department of Defense in March 2026 after the Pentagon designated Anthropic a "supply chain risk," a classification the company argues was used improperly to gain leverage in a contract dispute. A federal appeals court in Washington, D.C., heard oral arguments in that case in May 2026, with Anthropic's legal team arguing the department was "well equipped" to make procurement decisions without resorting to a public blacklisting of an American company.

Taken together, the closest thing to an "OpenAI vs. Claude legal battle" remains the 2025 API access dispute, a contractual and reputational conflict rather than a formal lawsuit, while each company's most significant current litigation runs on separate tracks: Anthropic against authors and, separately, against the Department of Defense, and OpenAI against publishers and authors including The New York Times. Industry observers continue to point to episodes like the Claude API cutoff as evidence of the increasingly adversarial posture major AI labs are taking toward one another as competition intensifies over model performance, enterprise customers and government contracts, even without a formal courtroom confrontation directly between OpenAI and Anthropic to date.