DuckDuckGo Installs Jump 33% as Users Flee Google's AI Search Overhaul

Google spent two weeks at I/O 2026 showing the world the future of search. The future, as Google sees it, is AI-first: expanded AI Overviews, a redesigned intelligent search box, and AI Mode baked into every query. Millions of dollars in engineering and marketing, all pointing toward a single vision of search where AI answers are the default and traditional blue links are the fallback.

Then users did something Google did not expect. They left.

DuckDuckGo reported a 33% week-over-week increase in US iOS installs in the days following Google I/O. Visits to its "No AI" search variant at noai.duckduckgo.com jumped 27.7%. These are not massive numbers in absolute terms. DuckDuckGo's market share hovers around 2-3%. But they are directionally unmistakable: a measurable segment of users is actively rejecting AI-first search and seeking alternatives that still feel like search.

This is the first quantifiable data point on user pushback against the AI search transformation. And it raises a question that the entire AI search industry needs to reckon with: what happens when you build the most powerful search engine in history and a meaningful slice of your users decides they do not want it?

What Happened at Google I/O

Google I/O 2026 was, by most accounts, the most significant Search redesign in the company's 25-year history. The announcements came fast and heavy.

AI Mode, previously an experimental feature in Search Labs, became a default experience for all US users. The traditional search box was redesigned as an "intelligent search box" that anticipates queries, suggests follow-ups, and surfaces AI-generated answers before the user finishes typing. AI Overviews expanded to cover even more query types, including local searches, product comparisons, and how-to queries.

The message was clear: Google Search is now an AI product. Traditional search still exists, but it is no longer the star of the show.

For many users, this was exciting. AI-generated answers save time. They synthesize information from multiple sources. They handle complex, multi-step queries that traditional search struggles with. But for a vocal and growing segment, the overhaul felt like a forced migration. The familiar blue links they had used for decades were being pushed aside. AI answers they did not ask for were appearing at the top of every query. And the simple, fast search experience they relied on was getting buried under layers of AI-generated content.

The DuckDuckGo Signal

DuckDuckGo's data, shared exclusively with The Verge, provides the first empirical evidence that this frustration is translating into behavior change.

The 33% increase in iOS installs is particularly telling. iOS users tend to be locked into Safari and Google as their default search engine. Changing search engines on iOS requires going into Settings, finding the Safari preferences, and manually switching the default. That is not a casual action. It requires intentionality. And a third more people made that intentional choice in the week after Google I/O than in the week before.

The 27.7% jump in visits to noai.duckduckgo.com is equally significant. This is DuckDuckGo's explicitly non-AI search variant, launched earlier in 2026 as a response to user demand for a search experience with zero AI features. No AI answers, no AI summaries, no chat interfaces. Just links. The fact that visits to this stripped-down search product surged immediately after Google's AI-heavy I/O suggests that the demand for non-AI search is not theoretical. It is real and growing.

DuckDuckGo CEO Gabriel Weinberg framed the data as evidence of a bifurcating market. "Some users want AI-powered search, and some users want traditional search," he told The Verge. "We are seeing growth in both, but the No AI growth is accelerating faster than anything we have seen in recent memory."

Context: Google's AI Search Stumbles

The DuckDuckGo data did not emerge in a vacuum. Google's AI search rollout has been accompanied by a series of high-profile stumbles that eroded user trust.

The most damaging was the AI Overviews "disregard" bug, first reported in late May 2026. Users discovered that AI Overviews were generating bizarre and incorrect answers when queries contained certain action verbs. Questions like "how to disregard [X]" or "ways to ignore [X]" triggered AI-generated responses that literally disregarded the core premise of the query, producing answers that were technically responsive but substantively wrong. The bug went viral on social media, with users sharing screenshots of increasingly absurd AI Overview responses.

Around the same time, Google announced a crackdown on "back-button hijacking," a publisher practice where sites prevent users from returning to search results by trapping them on the page. While the crackdown was framed as a user-experience improvement, it also highlighted the extent to which publishers were desperately trying to hold onto traffic that Google's AI Overviews were siphoning away.

Together, these incidents painted a picture of an AI search transition that was moving faster than Google's quality controls and faster than many users were comfortable with.

What This Means for the Search Market

The DuckDuckGo data points to a market that is fragmenting, not consolidating. The dominant narrative in AI search has been convergence: everyone is moving toward AI-first search, and the only question is who gets there first. But the data suggests a more complex reality.

AI-native search engines like ChatGPT Search and Perplexity are growing rapidly. ChatGPT has surpassed 600 million monthly active users, many of whom use it as their primary search tool. Perplexity has carved out a loyal niche among researchers and knowledge workers. These products are winning users who want more AI in their search, not less.

But simultaneously, DuckDuckGo's No AI variant is growing. Brave Search, which also offers a non-AI search mode, has reported similar trends. And the 33% iOS install spike suggests that some users are not switching to AI search alternatives but away from them.

This is not a zero-sum game. The search market is splitting into segments: users who want AI-powered search, users who want traditional search, and users who want both but on their own terms. Google's strategy of forcing AI search on everyone may be optimizing for the first group at the expense of the second and third.

Implications for Brands and Marketers

For brands, the fragmentation of search behavior has immediate strategic implications.

If you are only optimizing for Google's AI Overviews, you are reaching a shrinking share of total search attention. Some of your potential customers are using ChatGPT. Some are using Perplexity. Some are using DuckDuckGo's No AI variant, where traditional SEO still matters. And some are bouncing between all of these depending on the query and the context.

AI visibility strategy needs to account for multiple discovery surfaces. This means:

Do not abandon traditional SEO. Blue links still drive traffic, especially on non-Google search engines and for users who skip AI Overviews. The basics of title optimization, meta descriptions, and page speed still matter.

Optimize for AI citation across multiple platforms. Each AI search engine has its own source selection logic. ChatGPT favors content from well-known publishers and sites with clear entity signals. Perplexity prioritizes recently updated content with strong attribution. Google's AI Overviews lean on knowledge graph data and structured markup. A comprehensive AI visibility strategy addresses all three.

Monitor your AI visibility across surfaces. You cannot optimize what you do not measure. Use tools that track your brand's appearance in AI-generated answers across ChatGPT, Perplexity, Google AI Overviews, and Bing Copilot.

Prepare for further fragmentation. The search market is not done changing. Amazon's new Alexa for Shopping integration means product discovery is fragmenting into agentic commerce channels. Social platforms are building their own AI search features. The list of discovery surfaces will keep growing.

The Bigger Picture: Trust and Control

The DuckDuckGo data is ultimately a story about trust and control. Google built AI search features that many users find useful. But it made those features the default without giving users a clear way to opt out. The message, intentional or not, was that Google knows what users want better than users do.

That is a risky posture. Search is one of the most personal technologies people use. It is how people find information, make decisions, and navigate the world. When you change something that fundamental without user consent, you invite backlash. And when that backlash manifests as a 33% spike in competitor installs, it is not just a PR problem. It is a market signal.

Google is unlikely to reverse course. The AI search transition is central to its product strategy and its business model. AI Overviews create new ad inventory. AI Mode increases engagement time. The intelligent search box generates more query data. Every AI feature is designed to make search more valuable to advertisers and more difficult to leave.

But the DuckDuckGo data shows that some users are leaving anyway. And in a market where switching costs are near zero, a 33% directional signal can become a trend very quickly.

What to Watch

Several signals will determine whether the DuckDuckGo bump is a blip or the beginning of a sustained shift.

First, does the install growth sustain over multiple weeks? A single week of data is suggestive but not conclusive. If DuckDuckGo reports continued growth through June, the trend is real.

Second, do other non-AI search engines report similar patterns? Brave Search, Ecosia, and even Bing's traditional search mode could corroborate the signal.

Third, does Google respond with user controls? A simple toggle to disable AI Overviews or switch to a traditional search mode would address the most vocal complaints. Google has resisted this so far, but sustained backlash could change the calculus.

Fourth, what happens to ChatGPT and Perplexity growth? If AI search engines and non-AI search engines are both growing at Google's expense, the market is truly fragmenting. If only AI search engines are growing, the DuckDuckGo data is more niche than it appears.

For now, the data is clear on one point: the assumption that all users want AI search is wrong. Some do. Some do not. And the ones who do not are finding alternatives faster than anyone expected.

---

Is your brand visible across all search surfaces, AI and traditional? Run a free AI visibility audit to find out where you appear and where you are missing.