Google reportedly preparing to drop AI Overview

The search giant may be reconsidering its generative answer block in search results because it reduces ad clicks and requires heavy compute.
Google is reportedly considering moving away from AI Overview, the generative answer feature that appears at the top of search results. The block summarizes information and answers queries using AI, but recent signals suggest the company may be rethinking its role in the core search experience.
One reason is advertising economics. AI Overview can satisfy a user’s intent before they click through to websites, which may reduce traffic to publishers and lower the value of search ads. For Google, whose search business is still deeply tied to advertising, that is a significant concern.
The other issue is compute cost. Generating AI answers for billions of queries requires large amounts of GPU or TPU capacity, energy, and infrastructure. Some reports point to capacity constraints as a possible driver behind the decision to scale back or remove the feature.
There is no confirmed announcement from Google, so the change may be gradual rather than abrupt. The company could limit AI Overview to fewer query types, make it less prominent, or replace it with a lighter form of AI assistance while preserving the traditional search results page.
Why it matters
AnalysisIf Google retreats from AI Overview, it would show that generative AI in mass consumer products is constrained by monetization and infrastructure, not just model quality. It could also affect how quickly AI reshapes search advertising and publisher traffic.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
NotebookLM and Audio Overview (Podcast from Summaries)
A free research tool from Google based on Gemini models. It strictly adheres to uploaded sources (no hallucinations) and can transform dull reports into engaging conversational audio podcasts.
Gemini Flash & Pro (Google Gemini)
A family of multimodal models from Google DeepMind that combines a record context window (up to 2 million tokens), extreme generation speed (over 150 tokens/sec), and native perception of video and audio.
Google receives first Gemma response from orbit
Google’s Suncatcher project launched a satellite with TPU and Gemma, marking the first LLM response sent from space.
Google Loses Nearly $500 Billion in One Day
Alphabet shares fell 12.6% after its quarterly report as investors worried about massive AI spending and slowing ad revenue.