Feb 10, 04:11 PM

OpenAI Probes DeepSeek for Data Misuse as Microsoft Flags Concerns, R1 Model Faces Security Issues

OpenAI has initiated discussions with U.S. government officials regarding its investigation into allegations that China's DeepSeek improperly utilized data from OpenAI's technology. OpenAI suspects that DeepSeek trained its AI models using data obtained without authorization from OpenAI’s API. Microsoft flagged potential data misuse, and Chris Lehane, OpenAI's chief global affairs officer, stated that evidence has been found and the investigation is ongoing. Lehane compared OpenAI’s data use to reading a library book, while describing DeepSeek’s actions as akin to reselling a library book under a new cover. The allegations have raised broader concerns about data security and ethical AI practices. DeepSeek, known for its advanced AI models and open-source claims, has also faced scrutiny over vulnerabilities in its R1 model, which reportedly has a 100% jailbreaking success rate, prompting some governments to ban its use in official systems.

#OpenAI #China #DeepSeek #Microsoft #Chris Lehane #Lehane

Written with ChatGPT (GPT-4o).

Sources

Additional media

Image #1 for story openai-probes-deepseek-data-misuse-microsoft-flags-concerns-r1-model-faces-dc1f1824