Donald Trump’s AI ‘czar’, David Sacks, has claimed there is “substantial evidence” that DeepSeek “distilled” OpenAI’s knowledge to train its own AI models.

In an interview on Fox News, Sacks, who failed to provide the “evidence” for these accusations, suggested that DeepSeek used responses from OpenAI models to train its own – calling into question the cost-cutting claims of DeepSeek.

Sacks said: “There’s a technique in AI called distillation. It’s when one model learns from another model.” 

“What happens is that the student model asks the parent model a lot of questions, just like a human would learn. But AIs can do this by asking millions of questions, and they can essentially mimic the reasoning process that they learned from the parent model, and they can suck the knowledge out of the parent model.” 

“There’s substantial evidence that what DeepSeat did here is they distilled the knowledge out of OpenAI’s models and I don’t think OpenAI is very happy about this.” 

Silicon Valley was left reeling after DeepSeek AI chatbot – which it claims was developed for as little as $6m – shot to the top of the Apple charts in the US, overtaking OpenAI’s ChatGPT.

Earlier this week, Bloomberg UK reported that OpenAI and Microsoft are investigating whether data output from OpenAI’s technology was obtained in an unauthorised manner by a group linked to the Chinese firm.

According to sources close to the matter, late last year Microsoft’s security researchers observed individuals they believe may be linked to DeepSeek removing data using the OpenAI application programming interface. 

Reuters also reports that the National Security Council, the body which reports to the US government about foreign policy and national security, is reviewing the implications of DeepSeek.

If Sacks’ claims are found to be true, the distillation of knowledge would both violate OpenAI’s terms of service but also call into question the financial claims of DeepSeek, as the company would effectively be leveraging the hundreds of millions OpenAI has invested in its AI technology.

“We have to be the leader in AI”

During his interview on Fox, Sacks also backed the US’ heavy investment in AI data centres and development – a sentiment shared by tech CEOs from Meta and Microsoft in the companies’ respective earnings calls on Wednesday (29 January).

This was after the value of tech stocks was slashed on Monday amid fears that the rise of DeepSeek’s AI model would crush the demand for chips and reduce the appetite for investment in the sector.

“I think that an area where America can continue to lead is in a build-out of this infrastructure and having the most advanced chips,” explained Sacks. 

“So I think it’s a little bit of an overreaction to say that America doesn’t need big AI data centres anymore. I think we need the build-out of these big AI data centres. I think President Trump has expressed support for them and I think we need to make it easier to build those data centres, to get them permitted and licensed and to get power generated for those data centres. If we scale the biggest data centres, it is still an advantage.”