Bing expands its AI generative search

Searching the web for complex answers has always felt like homework. You open ten tabs, dodge three video ads, and skim through fluff just to find one troubleshooting step. Microsoft is trying to turn that whole mess into a single, cohesive page.

Following its early look in July, Microsoft just rolled out the next phase of its generative search in Bing. This update specifically targets complex informational queries that usually require deep research.

Tackling the tricky questions

Instead of just returning traditional blue links, Bing now reads and synthesizes information across the web to create custom structured layouts.

Microsoft highlighted detailed how-to questions as the prime target. Queries like how to effectively run a one on one or how to remove background noise from podcast audio will now trigger rich generative overviews instead of generic link lists.

For now, the feature is rolling out in beta exclusively to users in the United States. If you want to take it for a spin, typing Bing generative search into the search bar brings up a carousel of interactive demo queries.

For other topics, users can tap the Deep search button on the results page. When active, you will see a small label under the search box noting that results were enhanced with Bing generative search.

Trading speed for accuracy

Microsoft was upfront about one notable drawback: it takes a bit of time to load.

The company explained that generative results only appear when the system is confident in their accuracy and relevancy. That behind-the-scenes verification adds friction, but it theoretically prevents embarrassing hallucinations.

To keep web publishers from revolting, Bing keeps citations and source links prominent throughout the generated summary. The idea is to let users verify claims while continuing to funnel traffic back to original creators.

Users can also rate the summaries with thumbs up or thumbs down icons at the top, or submit detailed feedback at the bottom of the page.

Why it matters

For everyday researchers and creators, generative search can legitimately cut down research time on complex workflows. But because it is still US-only, carries noticeable latency, and operates in beta, it is not something most people will use for daily quick queries just yet.

Across the industry, this is Microsoft answering Google AI Overviews head-on. Both tech giants are racing to redefine what a search engine actually is, transforming it from a directory into an answer engine. While Microsoft insists citations will protect the web ecosystem, the reality is that fewer clicks will inevitably reach source websites if the AI answers everything on page one.

We think Microsoft's decision to gate this behind informational queries and the Deep search trigger is smart. It avoids burning computing power on simple searches that do not need AI. The clear trade-off is speed; modern users are trained to expect instant results, and waiting for an AI to synthesize multiple pages could push impatient searchers right back to standard links.