OpenAI accused of double standards by users — details

OpenAI accuses Chinese AI startup DeepSeek of using the results of its models to train its own chatbot. At the same time, OpenAI itself has repeatedly been the subject of lawsuits for copyright infringement and data misuse.

This was reported by Interesting Engineering.

OpenAI is accused of double standards

According to the Financial Times, OpenAI suspects that DeepSeek is "distilling" knowledge from ChatGPT, which may violate the terms of service.

"The problem is that you take this data outside the platform and use it to create your own model," said a source close to OpenAI.

OpenAI CEO Sam Altman has previously admitted that it is impossible to train powerful AI models without using copyrighted material. Following the DeepSeek accusations, many social media users criticized OpenAI's position.

"It's funny how OpenAI, which has been consuming our data without permission, is now accusing DeepSeek of the same thing," wrote one Bluesky user.

Ed Zitron, a well-known critic of artificial intelligence, sarcastically noted that OpenAI, a company that built a business on stealing the entire Internet, is now complaining that DeepSeek may have trained its model on ChatGPT results.

AI experts also point out that "knowledge distillation" is a common practice in the research world.

"Startups and academics often use the results of commercial language models such as ChatGPT to train their own systems," explained Ritvik Gupta, a researcher at the University of California Berkeley.

The scandal erupted after DeepSeek unveiled its new R1 AI model, which requires significantly fewer resources than analogues from leading Silicon Valley companies.

DeepSeek claims to have achieved a breakthrough in artificial intelligence development at a significantly lower cost. This raised the question - is their success the result of their own research, or is it partly based on the use of OpenAI data?

"OpenAI has no moral right to talk about copyright," said one social media user, alluding to the numerous accusations against the company of using protected materials without permission.

Despite the fact that OpenAI is outraged by the potential use of its technology, the company itself has repeatedly been sued for alleged copyright infringement.

As a reminder, the Chinese AI startup DeepSeek has shaken up Silicon Valley. The company has released AI models that compete with and even outperform OpenAI's flagship solutions. However, the American company suspects that they were created using their data.

We also wrote that another version of artificial intelligence was presented in China. The new AI model from Alibaba managed to outperform DeepSeek-V3.