AI News, February 25th, 2025

Here are three notable AI technology developments from February 25, 2025:
- DeepSeek Accelerates Launch of Advanced AI Model R2Chinese AI startup DeepSeek is expediting the release of its new AI model, R2, following the success of its R1 model, which significantly disrupted global markets. The R2 model promises enhanced coding and multilingual reasoning capabilities. Founder Liang Wenfeng has steered DeepSeek towards a research-focused culture, diverging from traditional hierarchical structures in China's tech industry. Despite facing scrutiny and restrictions in Western countries due to privacy concerns, DeepSeek's models are now integrated into numerous Chinese government and corporate systems. Reuters
- Microsoft Cancels Data Center Leases Amid AI Investment UncertaintyMicrosoft has canceled certain data center leases in the United States, equivalent to about two large data centers, amid doubts over substantial investments in artificial intelligence. These cancellations coincide with the rise of China's DeepSeek, which has efficiently developed generative chatbot technology at a lower cost and faster rate compared to U.S. rivals, causing tech stock market disruptions. While Microsoft still plans to spend over $80 billion on AI data centers this fiscal year, analysts speculate that the decrease in leases may be linked to an oversupply situation or OpenAI's shift to Oracle for their $500 billion AI project, Stargate. New York Post
- UK Creative Industries Launch 'Make It Fair' Campaign Against Copyright Law ChangesThe UK's creative industries have launched the "Make It Fair" campaign to challenge proposed government changes to copyright laws. These changes could allow tech companies to use British creative content for AI models without permission or payment, threatening the survival of industries contributing over £120 billion annually to the economy. The campaign urges public support to ensure fair compensation and transparency in the use of creative works by AI firms. The Guardian