Google Rolls Out Gemini 3 Across Core Services, Putting New AI in Front of Billions

Tech titans have bundled new engines into core products before—Microsoft welded Internet Explorer into Windows in 1998, sparking antitrust action; Facebook embedded Graph Search in 2013, redrawing data-collection norms. Google’s auto-activation of Gemini echoes those playbooks, but at 2025’s AI scale the stakes are larger: real-time parsing of private mail, photos and video, not just URLs. The launch sits at the intersection of two long arcs: the century-long march toward ubiquitous machine mediation of knowledge (from 1920s radio to 1990s web to today’s LLMs) and the 2010s rise of surveillance capitalism where data is mined first, permissions asked later. If unchecked, default-on multimodal AI could normalize continuous algorithmic oversight as electricity did industrial labor—baking the erosion of privacy and human agency into everyday tools for decades. Conversely, backlash like the San Jose lawsuit may become the 2020s equivalent of the 1974 U.S. Privacy Act, forcing legal guardrails that shape AI’s trajectory for the next 100 years.