Microsoft and Mistral's Multi-Billion Euro AI Deal, DeepSeek Privacy Concerns, and US-China AI Talks

Microsoft and Mistral have entered a multi-billion euro agreement to enhance AI infrastructure in Europe, integrating Mistral's models into Microsoft's platforms. This deal aims to reduce reliance on non-European providers, supported by thousands of NVIDIA GPUs for AI training and deployment. Public DeepSeek conversations have surfaced in Google search results, raising privacy concerns despite no evidence of private account breaches. Additionally, the US and China plan to hold talks in September to discuss AI risks in military, cyber, and labor contexts, focusing on governance and export controls. Other stories include OpenAI's ad revenue challenges, a cyberattack on Suno AI affecting 55 million users, and Alphabet delaying the launch of Gemini 3.5 Pro due to investor concerns over AI spending. Microsoft also announces changes to its Meeting Insights feature.