AI goes global, AI training lawsuits & more (December 18, 2025)

Good morning,

AI keeps stacking layers, faster models, wider adoption, louder lawsuits, and deeper geopolitical stakes. Today’s update adds a bit more context to each move so you can see where the pressure is building.

Let’s dig in 👇

━━━

Want to engage in conversation about AI news? Join the Byte-Sized Reddit Community

🚀 Model Upgrades & Global Expansion

⚡ Google drops Gemini 3 Flash

Google releases Gemini 3 Flash, promising faster responses with lower compute costs. The model is positioned as a lightweight workhorse for real time apps and agents. Google is clearly optimizing for scale, not just benchmarks. This also puts pressure on rivals to justify heavier, more expensive models for everyday use.

🌍 OpenAI and Anthropic eye Europe

OpenAI, Anthropic said to eye larger European offices as enterprise demand rises. The move signals deeper regulatory engagement, not just sales expansion. Europe is becoming too big to serve remotely. Local presence may also help both firms navigate tightening EU AI rules.

🇮🇳 India becomes the largest AI adopter

India now largest market in AI model adoption, according to Bank of America. Adoption is being driven by startups, developers, and enterprise tooling. This shifts global AI gravity beyond just the US and China. It also creates a massive proving ground for low cost, high volume AI products.

⚖️ Chips, Courts & Control

🎨 Adobe hit with AI training lawsuit

Adobe hit with proposed class action over alleged misuse of authors’ work. The case targets how creative data was sourced and labeled. Expect more scrutiny on “ethically trained” marketing claims. Similar lawsuits could ripple across the entire creative AI ecosystem.

🇨🇳 China’s AI chip Manhattan Project

How China built its Manhattan Project style push for domestic AI chips. Massive state backing is accelerating local alternatives to Nvidia. This is long term decoupling in motion. The effort prioritizes resilience over short term performance gaps.

🧠 The quiet war over compute

The global race for AI compute is reshaping geopolitics and supply chains. Nations are treating chips like oil reserves. Access to compute is becoming a strategic asset, not a commodity. Control over data centers may soon matter as much as model quality.

🧰 Tools of the Day

→ Shadow, a cloud PC for heavy AI, dev, and creative workloads
→ Netlify, deploy, scale, and manage modern web apps effortlessly
→ Sliq, automate finance ops and workflows for fast growing teams

⚡ Quick Hits

→ Clemson becomes second SC university exploring OpenAI contracts
→ Amazon appoints longtime AWS exec to lead new AI org
→ AI companies quietly rebrand AGI language
→ JP Morgan’s AI adoption hits 50% of employees
→ UN warns AI could widen global inequality without safeguards

🧠 TLDR

Faster models are shipping, Europe and India are scaling faster, and chips are now geopolitical leverage. Legal pressure is mounting on training data practices while nations race to lock down compute. AI is no longer just software, it’s strategic infrastructure.

Cheers,
David