OpenAI Engages Government Officials in DeepSeek Investigation

OpenAI has engaged with government officials regarding its ongoing investigation into DeepSeek, a Chinese company accused of using data from OpenAI's API without authorization. Chris Lehane, OpenAI's chief global affairs officer, confirmed these discussions during a Bloomberg TV interview.

The investigation centers on allegations that DeepSeek trained its AI models using data obtained improperly from OpenAI's systems. This comes after security researchers at Microsoft alerted OpenAI about potential data exfiltration by groups linked to DeepSeek. OpenAI has been reviewing evidence of this unauthorized use, which is believed to involve a process known as distillation.

Lehane explained the difference between OpenAI's data usage practices and those allegedly employed by DeepSeek. He compared OpenAI's approach to learning from a library book, whereas DeepSeek's method was likened to rebranding a library book and selling it as new. This analogy highlights the concerns over replication and unauthorized data use that OpenAI is addressing in its probe.