Google expands AI overviews

Google is changing what its AI search is supposed to feel like. Not just a box that spits out a summary, but a guided path through the web — with more visible sources, more community voices, and more prompts to keep digging. The move landed on May 6, when Google said AI Mode and AI Overviews would start surfacing richer link context, including previews from public discussions, blogs, and subscribed news sources. The bet is simple: AI answers are more useful when they feel less sealed-off from the open web. (blog.google) ### What actually changed? Google bundled five updates together. AI responses can now end with suggested next reads, so you get a set of deeper articles instead of a dead-end summary. Links from publications you already subscribe to can be highlighted in AI Mode and AI Overviews. And Google is adding more descriptive link cards, so you can see what kind of source sits behind the answer before you click. (blog.google) ### Why add forums and community posts? Because people already do this manually. A lot of searches end with “Reddit” or some version of “what do real people think?” Google is now trying to absorb that behavior into the AI layer. Its new previews can point to public online discussions, social posts, and other firsthand sources, with extra labels like a creator name, (blog.google)p of human opinions. (blog.google) ### Why is that useful for shopping? Shopping is where this format makes immediate sense. A normal search can show product pages and review sites, but an AI response can also compare tradeoffs, narrow options, and keep the context of your constraints — small kitchen, low budget, easy cleanup, whatever. Google has been steering AI Mode toward exactly that kind of us(blog.google)es. (blog.google) ### So is this still “search”? That is the real tension. Classic Google search mostly ranked pages and let you decide. AI Mode is doing more interpretation up front — summarizing, selecting, and sequencing sources before you ever visit them. The new features try to soften that by making sources more visible, but they also make Google more of an editor. That is a bigger product shift than it first sounds. (blog.google) ### Does it solve the trust problem? Not fully. The upside of richer citations is obvious — you can inspect where claims are coming from. But the catch is that an AI system can still misread a source, overstate a conclusion, or treat a joke post like real advice. That risk is not theoretical. Google’s AI search has already had some very public failures, and critics keep pointing out that better-looking citations are not the same thing as reliable reasoning. (techcrunch.com) ### Why is Google doing this now? Because AI search is turning into a product category, not a side feature. Google is under pressure to prove that its version of AI search is both more helpful than ten blue links and more grounded than a chatbot that improvises. The company’s recent updates all point in the same direction — conversational search, deeper follow-ups, and tighter integration with the broader web rather than a standalone answer engine. (blog.google) ### What does this mean for publishers? More visibility in AI answers — maybe. Google is clearly trying to reassure publishers by highlighting subscription links and sending users toward “next” articles. But there is still an uncomfortable tradeoff. If the AI answer gets good enough, fewer people may click through at all. So publishers could become more cited while also becoming less visited. (blog.google) ### Bottom line Google is expanding AI search by making the web more explicit inside the answer. That is smart. It may also be necessary. But the whole strategy still rises or falls on one thing — whether users believe the AI is guiding them to better information, not just prettier mistakes.