OpenAI has warned US lawmakers that its Chinese rival DeepSeek (深度求索) is using unfair and increasingly sophisticated methods to extract results from leading US artificial intelligence (AI) models to train the next generation of its breakthrough R1 chatbot, a memo reviewed by Bloomberg News showed.
In the memo, sent on Thursday to the US House of Representatives Select Committee on China, OpenAI said that DeepSeek had used so-called distillation techniques as part of “ongoing efforts to free-ride on the capabilities developed by OpenAI and other US frontier labs.”
The company said it had detected “new, obfuscated methods” designed to evade OpenAI’s defenses against misuse of its models’ output.
Photo: AFP
OpenAI began privately raising concerns about the practice shortly after the R1 model’s release last year, when it opened a probe with partner Microsoft Corp into whether DeepSeek had obtained its data in an unauthorized manner, Bloomberg previously reported.
In distillation, one AI model relies on the output of another for training purposes to develop similar capabilities.
Distillation, largely tied to China and occasionally Russia, has persisted and become more sophisticated despite attempts to crack down on users who breach OpenAI’s terms of service, the company said in the memo, citing activity it has observed on its platform.
Since DeepSeek and many other Chinese models do not carry a monthly subscription, the prevalence of distillation could pose a business threat to US companies such as OpenAI and Anthropic PBC that have invested billions of dollars in AI infrastructure and charge a fee for their premium services. That imbalance risks eroding the US advantage over China in AI.
OpenAI also highlighted other national security risks raised by DeepSeek’s gains, including that its chatbot had censored results about topics considered controversial by the Chinese government such as Taiwan and the Tiananmen Square Massacre.
When capabilities are copied through distillation, OpenAI said, safeguards often fall to the wayside, enabling more widespread misuse of AI models in high-risk areas such as biology or chemistry.
US Representative John Moolenaar, the Republican chair of the House committee, in a statement on Thursday said “this is part of the CCP’s playbook: steal, copy, and kill,” referring to the Chinese Communist Party. “Chinese companies will continue to distill and exploit American AI models to their advantage, just like when they ripped off OpenAI to build DeepSeek.”
OpenAI’s memo to the House committee suggests that its efforts to block distillation have failed to eliminate the problem. The company said an internal review suggests that accounts associated with DeepSeek employees sought to circumvent existing guardrails by accessing models through third-party routers to mask their source.
DeepSeek employees have also developed code to access US AI models and obtain outputs in “programmatic ways,” OpenAI said.
It also pointed to networks of “unauthorized resellers of OpenAI’s services,” also designed to evade the company’s controls.
White House AI czar David Sacks has previously warned about Chinese distillation tactics, telling Fox News last year that DeepSeek was “squeezing more juice” out of older chips, while also citing “substantial evidence that what DeepSeek did here is they distilled the knowledge out of OpenAI’s models.”
OpenAI’s warning about distillation also comes as many in Washington remain concerned that access to advanced AI chips might also accelerate DeepSeek’s progress. At the end of last year, US President Donald Trump moved to ease chip restraints and allow Nvidia to sell its H200 processors to China, chips that are about 18 months behind the leading Blackwell versions.
MediaTek Inc (聯發科), the world’s biggest smartphone chip supplier, yesterday said it plans to double investment in data center-related technologies, including advanced packaging and high-speed interconnect technologies, to broaden the new business’ customer and service portfolios. The chip designer is redirecting its resources to data centers, mainly designing application-specific integrated circuits (ASIC) with artificial intelligence (AI) capabilities for cloud service providers. The data center business is forecast to lead growth in the next three years and become the company’s second-biggest revenue source, replacing chips used in smart devices, MediaTek president Joe Chen (陳冠州) told a media event in Taipei. “Three or four years
Until US President Donald Trump’s return a year ago, when the EU talked about cutting economic dependency on foreign powers — it was understood to mean China, but now Brussels has US tech in its sights. As Trump ramps up his threats — from strong-arming Europe on trade to pushing to seize Greenland — concern has grown that the unpredictable leader could, should he so wish, plunge the bloc into digital darkness. Since Trump’s Greenland climbdown, top officials have stepped up warnings that the EU is dangerously exposed to geopolitical shocks and must work toward strategic independence — in defense, energy and
Motorists ride past a mural along a street in Varanasi, India, yesterday.
For the second year in a row, a Brazilian movie has wowed international audiences and critics, securing multiple Oscar nominations and drawing fresh interest in the Latin American giant’s film industry. Experts say the success of The Secret Agent, which has won four Oscar nominations, a year after I Am Still Here won Brazil its first Oscar, is no fluke, with a bit of a push from the country’s political climate. “This is neither a coincidence nor a miracle. It is the result of a lot of work, consistent policies, and, of course, talent,” Ilda Santiago, director of the Rio International Film
RSS