Microsoft reportedly investigates whether DeepSeek improperly used OpenAI’s models

OpenAI and Microsoft are probing whether DeepSeek used distillation to help train its rival reasoning model.

Transcript

00:00DeepSeek has caused a bloodbath for Wall Street, with tech titans losing billions in market value.

00:04But now Microsoft is reportedly investigating whether the startup used OpenAI's API to train its models.

00:10It all started when DeepSeek released its latest AI model that was allegedly developed on a $6 million budget,

00:15which is just a fraction of the billions OpenAI, Meta, and other tech giants have invested to train and develop their models.

00:22The Chinese startup's founder, Liang Wenfeng, originally started his career as a hedge fund manager,

00:26who launched the model as a hobby after buying thousands of NVIDIA graphic processors.

00:31The prospect of low-cost models led to a market sell-off that cost NVIDIA $600 billion in market cap,

00:36the biggest one-day loss in history.

00:38The release of DeepSeek AI from a Chinese company should be a wake-up call for our industries.

00:46But some investors are still bullish on U.S. tech stocks.

00:48Wedbush's Dan Ives dismissed concerns over the sell-off while dubbing DeepSeek as the Timu of AI.

00:54Other investors believe the market will rebound and DeepSeek's success will spur more investment into AI.