The sensational success of the Chinese startup DeepSeek has raised strong suspicions from OpenAI and Microsoft, who believe that the eastern company may have gained illicit advantages by using unauthorized data generated by the technology of Sam Altman’s company.
According to the latest rumors, the two American companies have launched an investigation to verify if individuals linked to DeepSeek have extracted sensitive information through the OpenAI API. The news shook the markets on the morning of January 29, contributing to the collapse of technology stocks related to artificial intelligence.
Microsoft discovers anomalies in the use of OpenAI’s API: the accusation against DeepSeek
According to sources close to the matter, during the autumn, Microsoft’s security researchers reportedly identified unusual attempts to access a vast amount of data through the OpenAI API.
This interface allows developers and business customers to integrate the U.S. company’s intelligenza artificiale models into their applications, but only with specific licenses and in compliance with strict usage policies. Investigations suggest that DeepSeek may have circumvented these restrictions to obtain valuable information, violating the terms of service.
In the best case, the Chinese startup would have exploited a vulnerability in the data access mechanisms. In the worst case, it would be a true large-scale theft, with legal implications still to be clarified. Microsoft and OpenAI have already initiated actions to strengthen the protection of their intellectual property and prevent further unauthorized access.
DeepSeek and the new R1 model: a threat to US AI?
At the center of attention is the R1 model, recently released by DeepSeek. Presented as an open-source artificial intelligence capable of reasoning similarly to humans, R1 has aroused the interest and concern of industry operators, challenging the dominance of OpenAI, Google, and Meta.
According to its developers, the model surpasses rival technologies in several areas, including mathematical skills and general knowledge, with development costs significantly lower than the American counterparts.
The impact of DeepSeek was immediately felt in the financial markets: the value of major AI companies, including Nvidia, Microsoft, and Alphabet, experienced a drastic reduction. In just a few hours, there was a total loss of nearly one trillion dollars, a sign of widespread fear regarding the loss of the U.S. technological primacy in the field of artificial intelligence.
The role of the White House and the accusations of David Sacks
The Trump administration quickly took a stance on the matter. David Sacks, the technology chief for the White House, stated that there is “substantial evidence” that DeepSeek used OpenAI’s models to train its own technology.
In an interview with Fox News, Sacks mentioned the method of “distillation,” a technique through which an AI can learn from the results generated by another model to replicate its capabilities.
The accusations by Sacks reflect a growing fear in Washington: the rapid development of China in the AI sector could be favored by illicit practices that undermine the competitiveness of American companies.
If the evidence were to confirm the suspicions, the DeepSeek case could lead to a strong response from the U.S. government, with new restrictions on technological exports and reinforced security measures.
OpenAI raises defenses: “Protecting our intellectual property is a priority”
OpenAI has also expressed concern about the phenomenon of distillation by Chinese companies and has confirmed that it is engaged in preventive actions to protect its models.
“The companies based in the People’s Republic of China constantly seek to replicate the technologies of the leading U.S. artificial intelligence companies”,
internal sources within Sam Altman’s company have declared.
The company has also reiterated its commitment to collaborate with the United States government to defend American technological supremacy and counter any attempt of misappropriation of its artificial intelligence.
With the DeepSeek case in the spotlight, the battle for leadership in AI is becoming increasingly intense, with geopolitical and market implications that could redefine the entire sector in the coming years.