According to Wired, technology company Anthropic has revoked OpenAI's API access to the Claude language model after discovering that OpenAI uses Claude to compare performance with internal models - which are banned in Anthropic's commercial terms.
Previously, OpenAI connected Claude with internal tools to evaluate criteria such as encryption, content writing and safety. Anthropic representative confirmed that OpenAI used their programming tools in the development of GPT-1, which Anthropic considered a serious violation of service terms.
However, Anthropic still maintains a limited level of access to serve the purposes of model standardization assessment and safety research.
OpenAI called the use of the service industry standard, expressing disappointment with Anthropic's decision, saying the company is still allowed to use OpenAI's API.
The tensions between the two sides reflect increasingly fierce competition in the field of artificial intelligence, especially as companies seek to protect core technology from the risk of being exploited by opponents.
