OpenAI has made the bold claim that the Chinese rival Deepseek has used its database to train its AI chatbot. They have claimed that they have proper evidence of distillation that is potentially from the parent company of R1 AI.
On Monday, Deep Seek’s claim of successfully achieving comparable results with an economical chatbot as those having far greater monetary and computing resources roiled the economic markets.
Nividia saw the biggest single-day tumble in terms of stock and Microsoft (the biggest supporter of OpenAI) also saw a decrease in stock value. This was because the investors were unclear about the greater spending in the US AI industry.
Therefore, undoubtedly, Microsoft was investigating the fact that Deepseek has used its database for bot training (saving the R&D and computing costs). On Wednesday, Microsoft claimed that it has suspected some individuals downloading large amounts of data from OpenAI sources.
Hence, this activity is reported as a Distillation that Microsoft suspects to be coming from Deepseek. This violates OpenAI’s terms of service.
Consequences to OpenAI
This claim has landed OpenAI in hot waters about its approach to other’s intellectual property rights. Many highlighted a lawsuit Microsoft is facing from The New York Times and other media outlets, where they claim that OpenAI used their information and data without permission.