The Chinese startup continues to disrupt the technology sector with its ‘low-cost’ AI – Accusations against DeepSeek also emerge from the White House.
DeepSeek’s emergence causes market shockwaves
The dynamic entry of Chinese artificial intelligence startup DeepSeek into the industry caused turmoil in global markets on Monday, disrupting the technology sector and raising concerns among its competitors.
According to reports from Bloomberg, Microsoft Corp. and OpenAI are now investigating whether OpenAI’s technology was unlawfully obtained by a group linked to DeepSeek. Sources indicate that Microsoft’s security researchers observed individuals, believed to be connected to DeepSeek, extracting large volumes of data using OpenAI’s application programming interface (API) last autumn. Software developers can pay for API access to integrate OpenAI’s proprietary AI models into their own applications.
Microsoft, OpenAI’s key technology partner and largest investor, alerted OpenAI to this activity, sources said. They also suggested that such activity could violate OpenAI’s terms or indicate an attempt to bypass OpenAI’s data access limitations.
DeepSeek’s AI revolution
Earlier this month, DeepSeek released a new open-source AI model named R1, designed to mimic human thought processes. Essentially, the Chinese startup demonstrated a way to develop artificial intelligence without requiring the massive financial investments made by Wall Street giants, positioning itself as a formidable rival to OpenAI’s ChatGPT.
This potential threat to industry leaders triggered a sharp decline in AI-related technology stocks, including Microsoft, Nvidia Corp., Oracle Corp., and Google’s parent company, Alphabet Inc., on Monday. The market reaction wiped out nearly $1 trillion in total capitalisation.
White House allegations
David Sacks, President Donald Trump’s AI policy chief, stated on Tuesday that there is “substantial evidence” suggesting DeepSeek leveraged OpenAI’s models to aid in developing its own technology. In an interview with Fox News, Sacks described a technique known as distillation, where an AI model uses the outputs of another for training purposes to develop similar capabilities.
“There is significant evidence that what DeepSeek has done here is extract knowledge from OpenAI’s models, and I don’t think OpenAI is very happy about it,” Sacks said, without specifying the evidence.
OpenAI’s response
Responding indirectly to Sacks’ remarks, OpenAI issued a statement: “We are aware that companies based in the PRC — and others — continuously attempt to distil the models of leading US AI companies.” The company further stated: “As the leading AI developer, we take countermeasures to protect our IP, including a careful process to determine which marginal capabilities are included in released models. Moving forward, we believe it is extremely important to work closely with the US government to better safeguard the most advanced models from adversaries and competitors attempting to acquire American technology.”
Also read: Tech selloff deepens as DeepSeek triggers AI rethink
Source: ot.gr
