OpenAI, which until recently created the most popular artificial intelligence service ChatGPT, said that Chinese companies are «constantly» trying to use American competitors to improve their AI models.
OpenAI, which until recently created the most popular artificial intelligence service ChatGPT, said that Chinese companies are «constantly» trying to use American competitors to improve their AI models.
OpenAI said in a statement, Reuters and the Financial Times reported, that the company is committed to protecting its intellectual property and carefully determining what advanced capabilities to include in the AI models it releases. The company believes that close cooperation with the US government is extremely important in order to best protect «the most powerful models from efforts by adversaries and competitors to seize American technology.»
OpenAI told the FT it had found evidence of «distillation» that could come from DeepSeek, a technique used by developers to get better performance on simpler models by using the results of larger, more powerful models. This allows them to achieve similar results at a much lower cost.
Distillation is a common practice in the industry, but concerns have been raised that DeepSeek could use it to create its own competing model, which violates OpenAI’s terms of service, the publication noted. OpenAI prohibits «copying» any of its services or «using the results to develop models that compete with OpenAI.»
However, the company declined to provide the FT with further comment or detail its evidence.
«As a leading AI developer, we take countermeasures to protect our intellectual property, including a thorough process for incorporating advanced capabilities into released models. It is critical to work closely with the US government to best protect the most effective models from attempts by adversaries and competitors to seize American technology,» OpenAI said in an official statement.
The company made the statement after the US White House said it was assessing possible national security concerns related to Chinese startup DeepSeek.
David Sachs, the White House’s top AI and cryptocurrency expert, said there was «substantial evidence» that Chinese company DeepSeek had «scooped» knowledge from OpenAI’s AI models. Sachs likened the process to theft and suggested that DeepSeek used OpenAI’s model responses to train its own.
