Microsoft and OpenAI are investigating whether Chinese AI startup DeepSeek improperly accessed OpenAI’s proprietary data to develop its own AI model, raising concerns over AI security, intellectual property theft, and escalating global competition in artificial intelligence. (Source: Image by RR)

DeepSeek’s AI Model R1 Claims to Rival U.S. AI Leaders at a Fraction of the Cost

Microsoft and OpenAI are investigating whether a group linked to the Chinese AI startup DeepSeek improperly accessed OpenAI’s data using its API. Microsoft security researchers reportedly observed individuals associated with DeepSeek extracting large volumes of data from OpenAI’s API, potentially violating OpenAI’s terms of service. While developers can legally license OpenAI’s API for integration into their own applications, the investigation seeks to determine whether DeepSeek circumvented restrictions to obtain unauthorized access to proprietary AI models.

The scrutiny, as reported in archive.md, comes as DeepSeek recently released its open-source AI model, R1, which claims to rival or surpass leading AI products from U.S. companies like OpenAI, Google, and Meta at a significantly lower cost. The emergence of DeepSeek’s technology has rattled the AI industry, contributing to a dramatic market reaction that saw AI-related stocks, including Microsoft, Nvidia, Oracle, and Alphabet, lose nearly $1 trillion in value. The threat posed by DeepSeek’s advancements has fueled concerns about the security of U.S. AI intellectual property and the competitive edge of American AI firms.

David Sacks, the U.S. government’s AI czar, stated that there is “substantial evidence” DeepSeek used a technique known as distillation—where an AI model is trained using the outputs of another model—to replicate OpenAI’s technology. In a Fox News interview, Sacks asserted that OpenAI is likely unhappy with this development, though he did not provide specific proof of wrongdoing. OpenAI did not directly comment on the allegations but acknowledged that PRC-based companies, among others, frequently attempt to extract knowledge from U.S. AI firms’ models.

In response to the allegations, OpenAI emphasized that it actively implements countermeasures to protect its intellectual property and ensure that cutting-edge AI capabilities are safeguarded against adversarial threats. The company highlighted the importance of working closely with the U.S. government to secure advanced AI models from unauthorized access and misuse. As the investigation continues, the case underscores broader concerns about AI security, intellectual property protection, and the growing global competition in artificial intelligence development.

read more at archive.md