Google Restricts Gemini Access, Microsoft Plans AI Integration Event

Google is adjusting its model access for Gemini. Free personal Google accounts will now access the Flash-Lite model by default. Previously, these accounts used a larger model, often referred to as 'Gemini Pro'. Users seeking to retain access to the larger model must subscribe to Google One's AI Premium plan.

This shift moves higher-capability models behind a paid tier. The Flash-Lite model offers a smaller context window of 1 million tokens. Google states this model is more efficient for many common tasks. The move follows similar strategies by other major model developers, segmenting access based on subscription.

Concurrently, Microsoft is preparing for a dedicated AI event. Scheduled for May 20th, the event will focus on AI advancements within Windows and Surface products. CEO Satya Nadella, Yusuf Mehdi, and Pavan Davuluri are slated to present. This event will precede Microsoft's annual Build developer conference.

Anticipated announcements include further integration of Copilot into Windows. New AI-centric features for upcoming Surface devices are also expected. This indicates a continuing strategy of embedding generative AI capabilities deeply within Microsoft's core software and hardware offerings.

These developments highlight differing commercialisation strategies. Google is refining its free-to-paid model access. Microsoft is accelerating AI feature integration across its ecosystem. How will these distinct approaches shape user adoption and model development priorities?