Is consumer AI heading for a duopoly? (gabrielweinberg.com)

Google appears to be replaying its old search-to-browser monopoly playbook in consumer AI: in the past 30 days Gemini downloads rivaled ChatGPT’s and no other competitor is close. That concentration matters because modern consumer AI isn’t a standalone app — it rides on distribution (search results, in-product prompts, and browser integration) and on proprietary engagement and crawling signals that Google uniquely controls. With Chrome and Google Search dominant on mobile and many sites limiting crawler access, Google can push AI features to a massive, built-in user base faster than smaller rivals, creating a de facto duopoly with ChatGPT that’s hard to dislodge. The technical and policy implications are significant. Distribution and first-party engagement data, plus privileged web access, give incumbents scale and training advantages; browser-level AI features can further entrench them. Regulatory action so far has only scratched the surface, so non-regulatory responses matter: deeper partnerships and consolidation among search engines, browsers and AI firms (examples include DuckDuckGo’s tie-up with You.com) could pool indexing, distribution and privacy-centric UX to compete at scale. Without that or stronger oversight, the market risks less competition, slower innovation and weaker privacy protections as AI becomes woven into search and browsing workflows.