Forget keywords: YouTube wants you to search by talking to AI

For two decades, searching on YouTube has followed a familiar pattern: type a few keywords, scan thumbnails, and click through a sea of results.

Now, that search box is starting to talk back.

Google is testing a new conversational AI feature called Ask YouTube, an experiment that turns video search into something closer to a chatbot — allowing users to ask full questions, refine queries in follow-ups, and receive results that blend AI-generated summaries, short clips, long-form videos, and highlighted timestamps.

Get updated faster and for FREE: Download the Gulf News app now - simply click here.

The shift is subtle in design, but significant in what it suggests about the future of discovery on the world’s biggest video platform.

Instead of searching “best beginner DSLR camera”, users might ask: “What’s the best camera for travel vlogging under $1,000?” Ask YouTube responds with a text summary, curated video selections, and relevant clips pulled from longer videos — effectively compressing what once took multiple searches into a single conversational prompt. Follow-up questions can narrow the results without starting over.

The experiment is currently limited to U.S.-based YouTube Premium subscribers aged 18 and older, rolling out through YouTube Labs, the company’s testing ground for early AI features.

This is not happening in isolation.

Across Google’s ecosystem, search is steadily moving away from keywords and toward conversation. Google Search’s AI Mode has already introduced multi-step conversational search on the web, powered by Gemini, while Gmail, Docs and Android are increasingly layered with AI assistants that understand intent rather than exact commands. Ask YouTube extends that strategy into video — arguably Google’s richest content library.

For creators and publishers, the implications could be profound.

Today, YouTube discovery depends heavily on titles, thumbnails, metadata and watch time signals. A conversational search layer changes that dynamic. Videos may increasingly surface because AI understands their context — what’s said in them, what they explain, and where in the video the answer lives — rather than simply because they rank well for keywords.

In effect, YouTube search may become less about finding a video, and more about finding an answer inside a video.

That also raises familiar questions around AI-generated summaries, particularly accuracy and attribution. Similar AI-powered search products have occasionally surfaced factual errors or condensed information in ways that change nuance. Early hands-on testing by The Verge found Ask YouTube useful, but not flawless — a reminder that conversational search still carries the same reliability challenges facing AI search everywhere.