Key points:

• DeepSeek, a Chinese AI startup, is under investigation for potentially misusing OpenAI data.
• Microsoft and OpenAI's inquiry follows DeepSeek's release of its R1 AI model, which rivals U.S. counterparts.
• The incident has ignited concerns over national security and the future of AI competition between the U.S. and China.

DeepSeek, a burgeoning Chinese AI enterprise, has become the center of controversy as Microsoft Corp. and OpenAI delve into allegations that the startup may have improperly accessed data from OpenAI's technology to develop its new AI model, R1. This model has rapidly garnered attention for its purported capabilities, raising unsettling concerns in the U.S. tech industry and stock market. Microsoft’s security researchers noted potential exfiltration incidents linked to DeepSeek in the previous autumn, where significant amounts of data were allegedly pulled through the OpenAI application programming interface (API), which is paid for by developers seeking to incorporate OpenAI's advanced AI models into their applications, according to sources familiar with the investigation. This investigation comes alongside reports that DeepSeek claims to have developed R1 at a mere fraction of the cost expended by U.S. rivals, prompting fears of the startup's potential to disrupt the current AI landscape dominated by American firms such as OpenAI, Microsoft, Google, and Meta.

Since its launch, DeepSeek's AI model R1 has claimed to rival the capabilities of established leaders, catching the attention of market observers and sparking a response that wiped nearly $1 trillion off the market value of major AI-related stocks, including those of Microsoft and Nvidia. OpenAI's representatives acknowledged ongoing attempts by Chinese companies to utilize distillation techniques—where minor AI models extract knowledge from more comprehensive models—to replicate advanced technology developed by the U.S. The urgency is heightened by statements from White House officials indicating that DeepSeek’s rapid advancements represent a significant challenge in the AI arms race.

David Sacks, a key advisor in the Biden administration on AI matters, indicated that there is substantive evidence pointing toward DeepSeek's potential misappropriation of OpenAI's data. “I don't think OpenAI is very happy about this,” he commented, underscoring increasing scrutiny placed on AI companies regarding their data practices. In parallel, Howard Lutnick, President Trump’s nominee for Commerce Secretary, criticized Chinese firms for allegedly leveraging U.S. technology against its own interests, contributing to an atmosphere of apprehension concerning national security and foreign competition.

The implications of this situation are broad, as both companies are probing defensive measures to safeguard their intellectual property and respond to the competitive threat posed by DeepSeek’s cost-effective offerings. OpenAI is collaborating closely with the U.S. government in these efforts to uphold its innovative edge and maintain the integrity of its operations amidst fears that such distillation tactics could lead to Chinese companies gaining critical ground in AI development.

For more about this unfolding issue, see the original reports from South China Morning Post, CNN, Business Insider, Forbes, and more top news sources.