DeepSeek's AI Model Shakes Up Industry, Paris Hosts AI Action Summit

A new open-source artificial intelligence (AI) model developed by the Chinese startup DeepSeek is making waves in the global tech community, rivaling leading models in performance while significantly reducing costs. DeepSeek's AI assistant briefly surpassed ChatGPT as the top-rated free app in the U.S. Apple App Store, impacting the market values of major tech companies like Nvidia. DeepSeek claims its model matches ChatGPT’s performance in mathematics and coding but uses less memory, cutting costs. According to DeepSeek, training their AI assistant required less than $6 million in Nvidia’s computing power.

However, the rise of DeepSeek has faced scrutiny, with Taiwan and Australia banning its use on federal devices due to security concerns1. Despite these concerns, Sam Altman, CEO of OpenAI, acknowledged the model as "impressive" and welcomed the competition.

In other news, Paris is positioning itself as the global capital for AI, hosting the Artificial Intelligence (AI) Action Summit on February 10 and 11, 2025. Sciences Po is rallying researchers and students to address the critical questions of AI's role in democratic societies. The summit will address this question to political leaders gathered in Paris. France plans to announce major investments, including for new data centers.

The AI Action Summit includes events such as the “Participatory AI Governance - Research & Practice Symposium” and “AI for Economic Inclusion”. The International Telecommunication Union (ITU) will host the AI for Good Global Summit 2025 in Geneva, Switzerland, from July 8 to 11, focusing on AI's role in sustainable development and addressing challenges like safety, governance, and societal impacts.

The International AI Safety Report will be presented to global leaders at the AI Action Summit in Paris. The report consolidates scientific findings on the safety of general-purpose AI to foster a shared global understanding of AI risks and mitigation strategies.