Anthropic has blocked OpenAI’s access to its Claude API, citing breaches of its terms of service. This decision followed OpenAI’s use of custom APIs to integrate Claude for benchmarking before the launch of GPT-5. Anthropic spokesperson Christopher Nulty indicated that OpenAI misused Claude for tasks related to coding, creative writing, and sensitive evaluations, violating their rules against developing competing products. OpenAI defended its actions as common industry practice for safety and benchmarking purposes, expressing disappointment over the API cut, yet confirmed it wouldn’t restrict Anthropic’s access in return. This isn’t the first instance of restricted access; in June, Anthropic suspended Claude for coding tool Windsurf amid acquisition rumors. To manage demand, Anthropic is implementing weekly usage caps on its Claude Code service, citing recent outages due to high usage. These restrictions underscore a growing trend in the tech industry to safeguard intellectual property and maintain model exclusivity amid competition.
Source link
Share
Read more