Recently, it was reported that ByteDance violated the terms of service when using OpenAI technology to develop a large language model, resulting in the suspension of the account. In response, ByteDance said it was communicating with OpenAI to clarify the misunderstanding.
According to the person in charge of ByteDance, at the beginning of this year, the byte technology team began to explore large models, and some engineers applied GPT's API services to experimental project research of smaller models, but the model was only tested and not launched. After April, the practice has stopped. ByteDance emphasized that as early as April, it had put forward an internal requirement that the data generated by GPT should not be added to the training dataset, and took measures to ensure that API calls to GPT meet the requirements of the specification.
ByteDance reportedly relied on OpenAI's API at every stage of the "Seed Project," with employees making heavy use of its technology, often even reaching OpenAI's access caps. The company ordered the team to stop using GPT-generated text at any stage of model development. At present, major Internet companies are releasing self-developed large models, while ByteDance's large model process is relatively slow. Perhaps this is also because they want to use OpenAI to accelerate the progress of their own large models.
In order to protect its own information security, OpenAI has already laid the groundwork in its own service agreement, prohibiting anyone, including individual developers, from using OpenAI's large model to develop competing products.
At present, ByteDance said that it will conduct a comprehensive inspection again to ensure strict compliance with the terms of use of the relevant services. According to the data of business investigation, the legal representative of Beijing Douyin Information Service is Zhang Lidong, with a registered capital of 2000000 yuan and an establishment date of 2012-03-09. Artificial intelligence