Google Search Enters AI Mode — and the Internet May Never Look the Same
It feels like ChatGPT. It sounds like Google. But who’s really deciding what you see?
On May 20 at Google I/O 2025, Google announced the rollout of AI Mode — a sweeping shift in how search works. Instead of giving you a list of links, search now enters a chat-like experience, where you type a question and get a direct, conversational response — just like ChatGPT and other AI chatbots. You can follow up naturally, refining your query like you’re texting with a helpful assistant — all powered by Google’s Gemini AI.
This move builds on Google’s launch of AI Overviews, introduced at Google I/O 2024, which generate instant AI summaries at the top of traditional search results. The rollout of AI Overviews marked the beginning of AI-native search, and according to Google, users have responded positively. In core markets like the U.S. and India, AI Overviews have led to a 10%+ increase in usage for the types of queries that trigger them. People are asking longer, more complex questions, clicking more links per session, and reporting greater satisfaction with results.
Now, AI Mode takes that shift further. It turns search into an ongoing dialogue, where instead of browsing through blue links, you receive a fully composed response, complete with cited sources and suggested follow-ups. This is no longer the search experience we’ve known for two decades. It’s guided navigation, with Google deciding what’s useful, what’s relevant, and what gets surfaced.
Reinvent the Engine While It’s Still Running
In rolling out AI Mode, Google is signaling something deeper: it’s willing to disrupt itself before someone else does.
AI Mode enables Search to dive deeper into the web than a traditional search on Google, Google wrote in a blog post.
For a company that built its empire on ranking and linking, shifting to AI-generated answers is nothing short of cannibalizing its own product. But Google seems to know it has no choice.
On May 7, Apple's Senior Vice President of Services, Eddy Cue, testified during the U.S. Department of Justice's antitrust trial against Google, revealing that Apple is "actively looking at" integrating AI-powered search engines into Safari. This kind of native, assistant-style browsing experience could challenge Google's grip on where — and how — people seek information online.
So Google is making its move: reinventing Search while it's still dominant, hoping to control the future before the future redefines it.
The Ad Tension Beneath the Surface
Bold reinvention comes at a price — and the first thing on the line is ad revenue.
Google’s multi-billion-dollar advertising business is built on a simple mechanic: users click. They scan results, visit websites, linger, and trigger monetizable events. But AI Mode threatens to break that chain. If users get everything they need from an AI-written summary — fast, clean, self-contained — where do the ads go? How does Google monetize a search experience that no longer looks like search?
One possible answer: embed sponsored content directly into the sources cited by AI summaries. These would appear alongside helpful links, but subtly drive attention to monetized partners. In short, ads that look and feel like organic results.
But that opens up a new kind of risk:
Will users know which parts are editorial and which are paid?
Will trust in the product suffer if commercial interests quietly guide the answers?
The same frictionless experience that makes AI Mode so compelling may also strip away the visibility — and viability — of Google’s core business.
The Trust Problem: What Gets Surfaced, and Why?
Compared to ChatGPT, AI Mode does have an edge in sourcing. It often includes links to the websites it pulls from — a meaningful step toward transparency. But the deeper issue isn’t just whether Google cites sources — it’s which sources it surfaces, and why.
Are users being pointed to well-researched, high-quality content? Or are they seeing SEO-optimized pages that rank well within Google’s own algorithm? The more AI responses are stitched together from opaque criteria, the harder it is to rely on what’s being recommended — especially if advertising dollars have any say in the outcome.
This leads to a bigger trust question:
Can Google preserve the integrity of its results if commercial incentives are embedded in the answers themselves?
In a blog post, Google says AI Mode is rooted in its “core quality and ranking systems,” with Gemini’s reasoning capabilities layered on to improve factuality. When confidence is low, it falls back to traditional results. But even the company admits: this is an early-stage product — and responses may unintentionally reflect a persona or opinion.
That’s a precarious place to be — especially when AI Mode increasingly becomes the user’s only point of contact with the web.
In contrast, upstart rival Perplexity.ai is positioning itself as the trustworthy alternative.
That kind of clarity isn’t just a feature — it’s baked into the product itself. It stands in contrast to Google’s curated, commercially entangled approach, where it’s often harder to tell whether a source is surfaced because it’s credible, convenient, or commercially advantageous.
In the race to build the next interface of knowledge, trust won’t just be a feature — it will be the foundation. And whether Google can carry that trust into the AI era, or lose it behind a seamless screen, is the question that will define its future.
What’s worth watching:
Ad placement inside AI answers: Will summaries quietly include paid suggestions — and how clearly will they be labeled?
Web sustainability: If fewer users click through, how long will creators and publishers keep investing in content that fuels the AI?
Source transparency: Will Google prioritize credibility — or convenience? Who decides what counts as trustworthy in a world where the source is optional?
And perhaps most critically:
Can Google preserve user trust in a system where the incentives behind every answer may never be fully visible?
Disclaimer: The views expressed in this post are my own and do not represent the views of any organization I’m affiliated with. This content is for informational purposes only and should not be considered investment advice.