Microsoft has launched an investigation into Chinese startup DeepSeek, following allegations that the firm used ChatGPT to build its recently released R-1 model.
The investigation is coming after the firm’s researchers accused some individuals linked to the Chinese company of pulling large amounts of data from OpenAI’s application programming interface (API), a Financial Times report said.
Microsoft researchers report usage anomalies in OpenAI API
According to reports, Microsoft mentioned that its security researchers discovered unknown data harvests using OpenAI’s API. After the discovery, the researchers alerted the company, which in turn alerted OpenAI. The activity was termed illegal data scraping and a breach of the company’s terms of service.
AI startup DeepSeek released its R-1 model earlier this week, with the release now being at the center of Microsoft and OpenAI’s dispute. The issue stemmed from the release of the AI model, which carries out accurate performance with increased speed and lower costs compared to ChatGPT’’s. The model also outperformed other AI models, including United States firms X and Meta under various benchmarks.
Allegations trail the release of DeepSeek’s AI model
According to a recent release, the White House crypto and AI Czar, David Sacks, recently mentioned that there is compelling evidence that DeepSeek used OpenAI’s model to train its AI. The process known as distillation is when an AI uses the output of another AI to train its capabilities. “There’s substantial evidence that what DeepSeek did here is they distilled knowledge out of OpenAI models, and I don’t think OpenAI is very happy about this,” Sacks said.
In a recent post, OpenAI mentioned that AI firms in China continue to take the initiative from the AI firms in the United States. However, the firm noted that it didn’t outrightly accuse the firm of any wrongdoing. While Microsoft continues to carry out its investigations, US officials are also taking steps to address concerns. A recent United States Navy directive mentioned that its personnel are prohibited from engaging with DeepSeek. CNBC reported that it may be down to the fact that they suspect that the Chinese government may be using AI to digitally access sensitive information.