OpenAI Accuses DeepSeek of Knowledge Distillation: “Substantial Evidence”

OpenAI, the company behind ChatGPT, has accused its Chinese competitor, DeepSeek, of using its work to develop rival AI models at a significantly lower cost. The claims, first reported by Bloomberg and BBC, suggest that DeepSeek may have engaged in “knowledge distillation,” a process where an AI model extracts information from another to enhance its own capabilities.
The allegations have raised concerns among U.S. tech leaders and government officials. Microsoft, OpenAI’s key investor, is reportedly investigating whether OpenAI’s proprietary data was misused. The White House has also weighed in, with AI and crypto czar David Sacks stating there is “substantial evidence” supporting OpenAI’s claims. 
“One of the things you’re going to see over the next few months is our leading AI companies taking steps to try and prevent distillation,” Sacks added.
National Security Concerns and Ethical Debates
The U.S. government has already taken measures to protect its AI sector, including restricting China’s access to advanced semiconductor chips. At his confirmation hearing, Commerce Secretary nominee Howard Lutnick stressed that further action might be necessary to safeguard American AI companies from intellectual property theft.
However, some experts argue that AI development has always relied on building upon publicly available research. Naomi Haefner, assistant professor of technology management at the University of St. Gallen, questioned whether DeepSeek truly trained its models independently. 
“If OpenAI’s claims are accurate, the idea that DeepSeek achieved comparable results at a fraction of the cost may be misleading,” she said. AI Venture Partner Crystal van Oosterom echoed this sentiment, noting that many American and European AI advancements have historically been used by global competitors.
DeepSeek Faces Security Scrutiny Amid Cyberattacks
While DeepSeek denies wrongdoing, the Chinese AI firm is facing mounting scrutiny. The U.S. Navy has reportedly warned personnel against using DeepSeek’s apps due to “potential security and ethical concerns.” Data safety experts have also urged caution, noting that the app collects large amounts of personal data stored on servers in China.
At the same time, DeepSeek claims it has been the victim of cyberattacks. The company announced on Monday that it would temporarily limit new registrations due to “large-scale malicious attacks” targeting its platform. A banner on DeepSeek’s website warns users that registration delays may occur.
A Growing Divide in the AI Arms Race
The ongoing dispute highlights the growing divide between U.S. and Chinese AI development. While OpenAI and Microsoft push for stronger protections, questions remain about how intellectual property laws should apply to artificial intelligence. Whether DeepSeek’s advancements are a product of innovation or improper appropriation is still under investigation, but the case underscores the increasing tensions in the global AI race.
Learn more about AI policies and governance to understand how you can use AI responsibility and safeguard your data.
The post OpenAI Accuses DeepSeek of Knowledge Distillation: “Substantial Evidence” appeared first on eWEEK.