DeepSeek-linked group suspected in OpenAI data probe

DeepSeek-linked group suspected of obtaining OpenAI data through unauthorised means, prompting an investigation by Microsoft and OpenAI.

OpenAI’s weekly active users have surpassed 400 million, highlighting the growing demand for AI tools.

Microsoft and OpenAI are investigating whether a group linked to Chinese AI startup DeepSeek accessed OpenAI data without authorisation. Bloomberg News reported that Microsoft’s security team detected large-scale data transfers last autumn using OpenAI’s application programming interface (API).

Microsoft, OpenAI’s largest investor, flagged the suspicious activity to the AI firm. DeepSeek, a low-cost Chinese AI startup, gained attention after its AI assistant surpassed OpenAI’s ChatGPT on Apple’s App Store in the US, causing a selloff in tech stocks.

White House AI and crypto adviser David Sacks suggested DeepSeek may have stolen US intellectual property by extracting knowledge from OpenAI’s models. An OpenAI spokesperson acknowledged that foreign firms frequently attempt to replicate its technology and stressed the importance of government collaboration to protect advanced AI models.

Microsoft declined to comment on the matter, while DeepSeek was unavailable for a response. OpenAI stated it actively counters unauthorised attempts to replicate its technology but did not specifically name DeepSeek.