Microsoft and OpenAI are investigating potential misuse of the latter’s AI technology by a group allegedly linked to Chinese AI startup DeepSeek, sources revealed. The probe focuses on whether DeepSeek accessed OpenAI’s data without authorization, raising concerns over potential violations of OpenAI’s terms of service.
OpenAI vs. DeepSeek AI: What’s Going On?
According to sources, Microsoft’s security team detected unusual activity last fall, where individuals possibly connected to DeepSeek extracted a significant amount of data through OpenAI’s application programming interface (API). The API, which allows developers to integrate OpenAI’s AI models into their applications under a paid license, may have been exploited to bypass data restrictions. Microsoft, OpenAI’s key investor and technology partner, informed OpenAI about the activity.
The sources indicated that the individuals may have attempted to circumvent OpenAI’s limits on data access. If confirmed, such actions would breach OpenAI’s policies. One source stated, “Such activity could violate OpenAI’s terms of service or could indicate the group acted to remove OpenAI’s restrictions on how much data they could obtain.”
DeepSeek recently introduced R1, an open-source AI model designed to replicate human reasoning. The company claims that R1 performs at or above the level of leading AI models developed by OpenAI, Google, and Meta, particularly in mathematics and general knowledge. DeepSeek also emphasized that R1 was developed at a significantly lower cost than its US counterparts.
The release of R1 has intensified competition in the AI sector, affecting market sentiment. On Monday, stocks of AI-related companies, including Microsoft, Nvidia, Oracle, and Alphabet, saw sharp declines, collectively wiping out close to $1 trillion in market value.
Trump’s Crypto & AI Czar Comments
David Sacks, artificial intelligence czar under former President Donald Trump, weighed in on the matter, suggesting that DeepSeek might have leveraged OpenAI’s technology for its own development. Speaking to Fox News, he referenced a method called distillation, where an AI model learns from another’s outputs to build similar capabilities. Sacks stated there is “substantial evidence” that DeepSeek used OpenAI-generated data in this way.
The investigation remains ongoing, with OpenAI and Microsoft working to determine the extent of the Chinese AI firm’s potential misuse of proprietary AI outputs. Neither company has officially commented on any legal actions or enforcement measures that might follow.
Latest Articles
Sam Altman’s Viral Comment on AI Resurfaces Amid DeepSeek Disruption
NRF Invests $32M in AI Startup Harrison.ai to Improve Medical Diagnostics
Kore.ai Introduces AI for Process, Transforming Enterprise AI Adoption
What is your reaction?
