DeepSeek Under Scrutiny: U.S. Officials and Tech Giants Allege Data Misappropriation

TLDR

• Microsoft and OpenAI investigating potential unauthorized data access by DeepSeek-linked individuals through OpenAI’s API
• DeepSeek’s free AI assistant surpassed ChatGPT on Apple’s App Store, triggering $600B tech stock selloff
• U.S. officials, including Commerce Secretary nominee Howard Lutnick, accuse DeepSeek of IP theft
• U.S. Navy banned personnel from using DeepSeek’s technology citing security concerns
• Italian regulators gave DeepSeek 20 days to explain their data collection and storage practices

The artificial intelligence landscape faced fresh turbulence this week as Chinese startup DeepSeek found itself at the center of multiple controversies. Microsoft and OpenAI launched an investigation into suspected unauthorized data collection by individuals allegedly connected to DeepSeek, according to Bloomberg News.

The investigation stems from events in fall 2024, when Microsoft’s security researchers detected what they believed to be DeepSeek-linked individuals extracting large amounts of data through OpenAI’s API. This interface serves as the primary method for developers and business customers to access OpenAI’s services.

The situation gained additional attention when DeepSeek’s free AI assistant overtook OpenAI’s ChatGPT on Apple’s App Store in the United States. This development triggered a substantial market reaction, leading to a tech stock selloff that particularly affected industry giant Nvidia, which saw nearly $600 billion erased from its market value.

Howard Lutnick, speaking during his Commerce Secretary nomination hearing before the U.S. Senate, made direct accusations against DeepSeek. “They stole things, they broke in, they’ve taken our IP,” Lutnick stated, echoing broader concerns within the U.S. government about potential misappropriation of American technology by Chinese companies.

OpenAI is unhappy that DeepSeek trained on their data without consent or compensation.

Oh, the irony for all us authors unhappy with OpenAI for training on our data without consent or compensation! https://t.co/3BJYQHnYjz via @techcrunch

— Professor Toby Walsh FAA FTSE (@TobyWalsh) January 29, 2025

The White House AI and crypto czar, David Sacks, supported these allegations in a Fox News interview. Sacks pointed to what he called “substantial evidence” that DeepSeek had used distillation techniques to extract knowledge from OpenAI’s models, a practice that would violate OpenAI’s terms of service.

OpenAI responded to these developments with a measured statement, acknowledging ongoing attempts by China-based companies to replicate U.S. AI models. The company emphasized its commitment to protecting intellectual property through various counter-measures, including careful control over which capabilities are included in released models.

The U.S. Navy took preemptive action by directing its personnel to avoid using DeepSeek’s technology “in any capacity.” The military branch cited potential security and ethical concerns related to the model’s origin and usage in an email to troops.

Italian regulators joined the scrutiny, giving DeepSeek a 20-day deadline to provide detailed information about their data collection, storage practices, and AI training methods. This request came despite the information being publicly available in DeepSeek’s Privacy Policy.

The controversy sparked debate within the tech community. Perplexity CEO Aravind Srinivas offered a technical perspective, suggesting that DeepSeek’s success stemmed from innovative reinforcement learning approaches rather than simple imitation of existing models.

Bill Ackman, a prominent hedge fund manager, raised questions about potential market manipulation, suggesting that DeepSeek’s hedge fund affiliate might have profited from the market downturn through strategic options trading.

What are the chances that @deepseek_ai’s hedge fund affiliate made a fortune yesterday with short-dated puts on @nvidia, power companies, etc.? A fortune could have been made.

— Bill Ackman (@BillAckman) January 28, 2025

The tech community’s response included pointed commentary about the broader implications for the industry. Some observers noted the irony of OpenAI’s complaints about unauthorized data usage, given similar criticisms previously directed at their own training practices.

Critics like Edward Zitron framed the controversy as symptomatic of deeper issues within American tech companies, suggesting that accusations against DeepSeek might mask underlying problems with innovation and competitiveness in the U.S. tech sector.

DeepSeek has maintained silence throughout the controversy, declining to respond to multiple requests for comment from various news organizations.

The latest development saw OpenAI confirming their ongoing investigation into the matter, with a spokesperson stating they are “reviewing indications that DeepSeek may have inappropriately distilled our models.”