AI Update, January 31, 2025: AI News and Views From the Past Week

Catch up on the latest AI news and developments from the past week:

Google quietly releases Gemini 2.0 Pro Experimental, its newest flagship AI model, aimed at enhancing coding and mathematical capabilities. Mistral AI launches compact, high-performing AI model that rivals larger models. US Copyright Office allows AI-assisted works to receive copyright with human creativity.

Alibaba unveils Qwen 2.5 AI model, claiming it outperforms leading models. Convergence AI launches Proxy, a customizable AI assistant for automating digital tasks. Legal advisories outline how AI companies may be violating state laws. Society Library introduces AI-powered fact-checking agents. OpenAI releases ChatGPT Gov for US government agencies. Apple revamps AI team and plans major Siri upgrades. Samsung adds AI customization for Sketch to Image in Galaxy S25. Apple reportedly developing multiple smart glasses concepts. Google invests $120M in AI education. OpenAI faces new copyright lawsuit from Indian publishers. All about DeepSeek this week.

DeepSeek's AI models, developed at a fraction of the cost of competitors, are challenging the dominance of US tech giants. DeepSeek unveils Janus Pro, an open-source rival to DALL-E 3. DeepSeek limits registrations after large-scale cyberattack. DeepSeek surpasses ChatGPT as the most downloaded app. Researchers claim to have recreated DeepSeek's core AI for just $30. OpenAI investigates whether DeepSeek copied its models. DeepSeek chatbot scores low in accuracy compared to rivals. DeepSeek founder Liang Wenfeng emerges as key figure in AI innovation in China.