Overview

• Anthropic revoked OpenAI’s Claude API access on August 1 after determining its engineers used Claude Code in violation of its commercial terms.
• Under Anthropic’s terms of service, clients are prohibited from using Claude to build competing AI models, reverse engineer its services or train rival products.
• The move echoes a June decision to cut off coding startup Windsurf’s access amid similar competitive concerns.
• Anthropic said it will continue to provide API access for recognized benchmarking and safety evaluations, in line with industry norms.
• Industry reports say GPT-5 is expected to launch next week with “auto” and reasoning modes, intensifying OpenAI’s push to compare performance against competitor models.