The DeepSeek controversy has taken a new turn. After rumors about misleading claims regarding its actual processing power, the Chinese startup is now facing serious accusations. Yesterday, Bloomberg reported statements from Microsoft suggesting that DeepSeek may have stolen data and intellectual property from OpenAI.
Please follow us on Facebook and Twitter.
Microsoft, OpenAI’s largest investor, has reportedly assigned its security team to investigate how DeepSeek allegedly accessed OpenAI’s API to develop its own language model.
According to the source, Microsoft has evidence that a group linked to DeepSeek leaked a large amount of data from the OpenAI API last autumn. Although not stated directly, Microsoft is essentially accusing the Chinese startup of using this data to develop its own language model.
This issue has gained even more attention because DeepSeek has recently become the most popular AI in the world. In the United States, the app has surpassed ChatGPT in downloads, a trend that has also been seen in European countries.
US Supports DeepSeek Theft Theory
Adding to the controversy, following Bloomberg’s report, the White House Spokesperson for Artificial Intelligence and Cryptocurrencies, David Sacks, has weighed in on the issue. He stated that he believes DeepSeek may have stolen part of the United States’ intellectual property, specifically referring to the OpenAI API.
“There is substantial evidence that what DeepSeek did here was extract knowledge from OpenAI models,” Sacks said.
Meanwhile, an OpenAI spokesperson echoed Sacks’ remarks, suggesting—without directly naming DeepSeek—that Chinese companies are attempting to replicate language models developed by U.S. firms.
“We take steps to protect our intellectual property, including a careful process to determine which cutting-edge capabilities to include in fielded models. We believe that as we move forward, it is critically important to work closely with the U.S. government to better protect the most capable models from efforts by adversaries and competitors to steal U.S. technology,” the spokesperson said in remarks to Reuters.
Of course, DeepSeek has not commented on the matter yet, and it’s possible that it never will. In the U.S., there seems to be growing concern about the progress of Chinese AI, with numerous rumors and leaks emerging as an attempt to halt its success—or at least not give all the credit to engineers based in China.