Google's AI Overviews now quote Reddit posts and surface more source links
- Google updated AI Mode and AI Overviews on May 6 to quote Reddit and forum posts directly, while adding more visible links inside answers. - The new design adds creator names, handles, community labels, deeper reading suggestions, and even links from users’ own news subscriptions. - That matters because Google is answering more searches itself while insisting extra links can still send valuable traffic to the web.
Google Search is getting a little more like Reddit — and a little more like a chatbot with footnotes. On May 6, Google said AI Overviews and AI Mode will now pull in “perspectives” from public discussions, including Reddit and other forums, and show more links directly inside AI answers. The pitch is simple: people want firsthand advice, not just polished webpages. But the stakes are bigger than a UI tweak, because every extra thing Google answers on the page is one less reason to click away. (blog.google) ### What changed in Search? Google’s new update does two things at once. First, AI answers now surface more direct links in the body of the response instead of making users hunt for sources at the end. Second, those answers can include previews of social posts, forum threads, and other firsthand commentary — the kind of stuff (blog.google)community name so the source looks less anonymous before you click. (blog.google) ### Why Reddit specifically? Because people already use Google that way. For product advice, travel tips, symptoms, dating questions, or “what is this actually like,” users often trust messy firsthand accounts more than SEO pages. Google is basically admitting that behavior is part of modern search and formalizing it inside th(blog.google)er — public discussions, social media, and other firsthand sources. (theverge.com) ### What do the new links look like? This is the less flashy but maybe more important part. Google says many AI responses will now end with suggestions for where to go next — unique articles or deeper analysis on different angles of the topic. It is also highlighting links from a user’s own news subscriptions inside AI responses, (theverge.com)ally, Google is trying to make AI answers feel less like dead ends. (blog.google) ### Is this really about helping websites? Google says yes. The company has been arguing for months that AI Overviews show more links than before, drive more queries, and send “higher quality” clicks — meaning users who do click are more likely to stay. It also says overall organic click volume from Search has been relatively (blog.google)t is Google’s defense against the zero-click backlash. (blog.google) ### So why are publishers nervous? Because “higher quality clicks” is not the same thing as “more clicks.” Critics have been arguing that AI Overviews answer too much on the results page, cutting publishers out before a visit ever happens. That tension gets sharper when Google starts quoting the web more di(blog.google)und the answer, but Google keeps the user longer. (blog.google) ### What’s the reliability problem here? Reddit is useful, but it is also chaotic. Google’s AI search has already had some infamous failures involving sarcastic or bad forum advice getting treated too seriously. TechCrunch notes that this system has improved, but the underlying problem has not vanished: lar(blog.google)rce labels helps. It does not solve that. (techcrunch.com) ### Why does this matter beyond Google? Because this is where search is heading. The old model was ten blue links and your judgment. The new model is an AI summary that tries to pre-chew the web, then offers selected exits. Google is now making those exits more obvious and more human — creators, com(techcrunch.com)he web for you.” (blog.google) ### Bottom line Google is not backing away from AI answers. It is making them feel more sourced, more social, and more clickable. But that is still a version of search where Google does more of the reading first — and everyone else fights harder for the remaining clicks. (blog.google)