- OpenAI and Microsoft suspect that Chinese AI company DeepSeek used OpenAI’s data to develop its own competing models, potentially violating OpenAI’s terms of service.
- Microsoft security teams detected large-scale data extraction from OpenAI developer accounts in late 2024, which were allegedly linked to DeepSeek.
- The case raises concerns over AI intellectual property security, with OpenAI working closely with U.S. authorities to prevent unauthorized replication of its technology.
Silicon Valley faces fresh concerns over intellectual property security after OpenAI and Microsoft launched an investigation into whether China-based AI company DeepSeek leveraged OpenAI’s models to develop its own AI systems. The probe follows suspicions that DeepSeek, which has gained attention for its cost-effective AI models rivaling OpenAI’s flagship offerings, may have accessed OpenAI’s data improperly.
According to sources, Microsoft security teams detected large amounts of data being extracted through OpenAI developer accounts in late 2024. These accounts, allegedly linked to DeepSeek, may have facilitated the unauthorized use of OpenAI’s artificial intelligence models. If confirmed, this would represent a breach of OpenAI’s terms of service, which allow integration of its AI via API but prohibit the practice of distillation—training new models by extracting knowledge from existing ones.
Distillation is a common technique in AI development that allows companies to train smaller, cost-effective models by learning from more advanced systems. While OpenAI itself has relied on massive datasets scraped from the internet to train its models, it maintains strict controls over how its own AI-generated outputs can be used. OpenAI has not disclosed the specifics of the evidence it gathered but asserts that DeepSeek’s use of distillation may have crossed ethical and legal boundaries.
The case highlights growing tensions over AI technology, particularly between the United States and China. With artificial intelligence playing an increasingly strategic role in global competition, concerns over data security and intellectual property theft have become more prominent. The U.S. government has previously warned about Chinese companies attempting to extract sensitive AI knowledge from American firms, prompting OpenAI to work closely with authorities to safeguard its technology.
As AI development accelerates, industry leaders and governments are expected to tighten security measures to prevent unauthorized model replication. The outcome of OpenAI’s investigation could set a precedent for how companies protect their intellectual property in an era where AI training methods are becoming increasingly sophisticated and difficult to regulate.