Archive — US — techcrunch
OpenAI spoke to government officials about its DeepSeek probe
OpenAI has engaged with government officials regarding its investigation into DeepSeek, alleging the company improperly trained its AI models using data from OpenAI’s API. OpenAI's chief global affairs officer, Chris Lehane, highlighted differences between OpenAI's training methods and those of DeepSeek, amid criticism of OpenAI's own copyright practices.
Key points
- OpenAI claims DeepSeek misused data from its API for AI training.
- Chris Lehane has stated that OpenAI's training resembles learning from a book, contrasting with DeepSeek's methods.
- DeepSeek's practices have drawn criticism similar to accusations faced by OpenAI from various publishers.
- The investigation raises questions about copyright practices in AI model training.
Why it matters
- AI copyright issues have surged as companies increasingly train models on diverse data sources.
- Similar lawsuits in the tech industry highlight ongoing debates about data usage and intellectual property.
- By 2025, trends in AI development may focus more on ethical sourcing of training data and compliance with copyright laws.
The ongoing scrutiny of both OpenAI's and DeepSeek's practices reflects a critical examination of data usage in AI, a significant issue as the industry evolves towards greater ethical considerations and regulatory compliance.