Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Microsoft probing if DeepSeek improperly obtained OpenAI data

Jan 29, 2025 - financialpost.com
Microsoft and OpenAI are investigating whether DeepSeek, a Chinese AI startup, improperly obtained data from OpenAI's models. Microsoft security researchers observed potential data exfiltration via OpenAI's API, which could violate OpenAI's terms of service. DeepSeek recently released an AI model, R1, which claims to rival leading U.S. products at a lower cost, causing a significant drop in AI-related tech stocks. David Sacks, Trump's AI czar, suggested that DeepSeek used a technique called distillation to develop its technology by leveraging OpenAI's outputs.

OpenAI has not directly addressed these allegations but acknowledged ongoing attempts by companies, including those from China, to distill models from leading U.S. AI firms. OpenAI emphasized the importance of working with the U.S. government to protect its intellectual property. DeepSeek claims its R1 model was developed using open-source systems, contrasting with OpenAI's closed systems. The situation highlights ongoing debates about open-sourcing AI technology and its implications for industry competition.

Key takeaways:

  • Microsoft and OpenAI are investigating whether DeepSeek improperly obtained data from OpenAI's technology.
  • DeepSeek released an AI model called R1, which rivals leading U.S. products and was developed at a lower cost.
  • The potential threat from DeepSeek caused a significant drop in market value for major tech stocks tied to AI.
  • OpenAI is taking countermeasures to protect its intellectual property from unauthorized use by competitors.
View Full Article

Comments (0)

Be the first to comment!