OpenAI, the company that created ChatGPT, has expressed dissatisfaction over competitors, notably those in China, utilizing its work to quickly expand their own artificial intelligence (AI) capabilities.
This week, the abrupt appearance of DeepSeek, a Chinese program that can mimic ChatGPT’s performance, ostensibly for a fraction of the price, severely damaged OpenAI’s and other US companies’ standing as global leaders in AI.
According to Bloomberg, Microsoft is looking into the unauthorized use of data owned by OpenAI, in which it has a significant stake.
DeepSeek and Microsoft have been contacted by the BBC for comment.
David Sacks, the newly designated “AI and crypto czar” in the White House, has expressed similar worries to those of OpenAI.
In an interview with Fox News, he hypothesized that DeepSeek would have improved by using OpenAI’s models—a technique known as knowledge distillation.
“There’s substantial evidence that what DeepSeek did here is they distilled the knowledge out of OpenAI’s models,” said Sacks.
“I believe that one of the things that will happen in the coming months is that our top AI businesses will try to stop distillation. Some of these imitation models would undoubtedly slow down as a result.
In the guise of national security, the US has already taken action to protect its AI advancements by enacting regulations that aim to deny China access to cutting-edge technology and direct investments to the US.
Howard Lutnick, Trump’s nominee for Commerce Secretary, expressed similar worries about theft during his confirmation hearing on Thursday and hinted at the possibility of additional US intervention to safeguard US AI companies.
“What this showed is that our export controls, not backed by tariffs, are like a whack-a-mole model,” said Lutnick.
Chinese and other businesses were “constantly trying to distil the models of leading US AI companies,” according to a statement released by OpenAI.
“As we go forward… it is critically important that we are working closely with the US government to best protect the most capable models,” it stated.
What is DeepSeek?
Is using DeepSeek safe?
How do ChatGPT and DeepSeek compare? China’s DeepSeek upends the AI sector.
Related Post-Users claim that the Chinese AI chatbot DeepSeek censors itself in real time.
“Deceptive” statements
The subject of distillation, according to Naomi Haefner, assistant professor of technology management at the University of St. Gallen in Switzerland, could cast doubt on the idea that DeepSeek developed its product for a quarter of the price.
“It is unclear whether DeepSeek really trained its models from scratch,” she stated.
“OpenAI has expressed their suspicion that DeepSeek may have stolen a significant quantity of their data.
“If this is true, then the assertions that the model was trained at a very low cost are false. We won’t be certain whether such cost-effective training is indeed feasible until someone duplicates the teaching methodology.
“DeepSeek has clearly built upon publicly available research from major American and European institutions and companies,” concurred Crystal van Oosterom, AI Venture Partner at OpenOcean.
The issue with the notion of “building on” other people’s work is unclear, though.
This is particularly true in the field of artificial intelligence (AI), where prominent US AI companies have often been accused of violating intellectual property rights.
Ethics and security
According to White House press secretary Karoline Leavitt, US officials are also thinking about the ramifications of DeepSeek’s rise for national security.
“I spoke with [the National Security Council] this morning, they are looking into what [the national security implications] may be,” Ms. Leavitt said, echoing earlier statements made by US President Donald Trump that DeepSeek should serve as a wake-up call for the US tech sector.
The US Navy reportedly prohibited its members from using DeepSeek’s apps because of “potential security and ethical concerns” before making the announcement.
The US Navy has warned its employees via email not to use the DeepSeek app because of “potential security and ethical concerns associated with the model’s origin and usage,” according to CNBC.
A BBC News request for comment was not immediately answered by the Navy. Experts in data security have cautioned users against using the program because it gathers a lot of personal information and keeps it on Chinese servers.
DeepSeek claims that it has been the victim of cyberattacks in the interim. It announced on Monday that “large-scale malicious attacks” on its software would compel it to temporarily restrict registrations.
The company’s website displays a notice stating that due to the attacks, registration may be busy.