The State of AI Search in 2026: Every Stat You Need

Every stat in this post links to its source. If a number doesn’t have a link, it shouldn’t be here. We’ve pulled from Backlinko, Semrush, Grand View Research, McKinsey, Gartner, Pew Research, and others. Where data conflicts (and it does), we’ve noted both figures.

Here’s where AI search stands in early 2026.

Market size: how big is AI search right now?

The global AI market is worth an estimated $900 billion in 2026, according to Precedence Research. That’s up from $757.58 billion in 2025.

The generative AI segment specifically (including search) sits at $22.21 billion as of 2025, with Grand View Research projecting it will hit $324.68 billion by 2033, growing at a compound annual growth rate of 40.8%.

For context on investment velocity: VC funding into AI startups hit $45 billion across more than 1,000 deals in Q3 2025 alone, according to Crunchbase.

Enterprise spending on generative AI solutions (software, hardware, and IT services combined) is projected to reach $143 billion by 2027, up from $16 billion in 2023, per eMarketer. That’s nearly a 9x increase in four years.

Spending on generative AI software alone is estimated to reach $52.2 billion by 2028, per S&P Global. It was $5.1 billion in 2023.

Platform adoption: who’s using what

ChatGPT

ChatGPT has over 900 million weekly active users as of December 2025. That’s a 3x increase from 300 million in December 2024.

The growth curve has been relentless:

• November 2023: 100 million weekly users
• August 2024: 200 million
• December 2024: 300 million
• February 2025: 400 million
• August 2025: 700 million
• December 2025: 900 million

(Sources: Reuters, The Information, TechCrunch)

ChatGPT.com gets roughly 5.6 billion monthly visits as of late 2025. It ranks as the fourth most visited website in the world.

ChatGPT Plus has over 10 million paying subscribers. Enterprise, Team, and Edu plans add another 5 million paid users, according to CNBC.

OpenAI’s monthly revenue reached $1 billion in October 2025. The company expects $12 billion in total 2025 revenue.

ChatGPT holds 62.5% of the B2C AI subscription market. No other paid AI tool has more than 6%.

Perplexity

Perplexity processes over 20 million daily queries. That sounds small next to Google’s 8.5 billion daily searches. But Perplexity queries are fundamentally different: they’re conversational, research-oriented, and tend to include brand consideration signals. A Perplexity query about “best accounting software for law firms” is closer to a buying decision than a Google search for “what is accounting software.”

Perplexity’s citation patterns lean heavily on Reddit and community sources. Detailed.com found Reddit makes up 46.7% of Perplexity’s top 10 most-cited domains, compared to 4.3% for ChatGPT. If your brand lives on Reddit, you’re in a strong position on Perplexity. If it doesn’t, you’re invisible there.

Google AI Overviews

Google AI Overviews now trigger on roughly 16% of all Google searches, per Backlinko’s AI Overviews research. Semrush studied 200,000 keywords and found 80% of AI Overviews on desktop target informational intent queries.

The average AI Overview is 119 words on desktop and 91 on mobile. They contain an average of 4.7 source links, with some containing up to 302 links in complex travel and research queries.

Health, People & Society, and Science categories trigger AI Overviews most frequently on desktop, according to Semrush Sensor data.

Claude

Anthropic doesn’t publish user numbers the same way OpenAI does, but the Apple partnership gives Claude a distribution advantage most AI companies would kill for. Claude is integrated into Apple Intelligence across iPhone, iPad, and Mac, which means billions of potential touchpoints even for people who’ve never heard of Anthropic.

Gemini

Google’s Gemini is embedded across Google Search, Gmail, Docs, and Android. As a standalone product it trails ChatGPT, but as an embedded AI layer across Google’s product suite, it reaches billions of users passively. The distinction matters: Gemini’s brand recommendation surface isn’t just gemini.google.com. It’s everywhere Google is.

How AI search is changing traffic patterns

The biggest shift isn’t that people are using AI search. It’s that AI search is eating the clicks that used to go to websites.

Gartner predicted in February 2024 that traditional search engine volume would drop 25% by 2026 due to AI chatbots and virtual agents. We’re tracking toward that number.

Ahrefs found that position-one click-through rates for informational keywords dropped from 7.6% in December 2023 to 3.9% in December 2025. That’s a 49% decline in two years. Half the clicks gone. Not because rankings changed, but because AI Overviews answered the question before anyone needed to click.

Here’s the counter-argument that makes this more interesting: AI search visitors convert 4.4 times better than traditional organic visitors. Fewer clicks, but the clicks that do come through carry more intent. Someone who gets a partial answer from AI and still clicks to your site is deeply interested. The window shopper traffic drops. The buyer traffic stays or increases.

That’s a painful trade-off for businesses that lived on informational traffic for ad revenue. And a tailwind for businesses that actually sell things.

Enterprise adoption: who’s using AI for search internally

McKinsey’s State of AI report found that 88% of global companies use AI in at least one business function, up from 78% the previous year.

In the US, 116.9 million people use generative AI at least monthly, per eMarketer. That’s roughly 35% of the US population.

54% of US companies with $500M+ annual revenue have already implemented generative AI in some areas, according to PwC.

The National Bureau of Economic Research found 27.1% of American working adults use generative AI for work. Not personal. Work. Meaning a quarter of the workforce is already using AI to find answers, research products, and evaluate vendors during business hours.

What people actually use AI search for

Exploding Topics research found the top three AI usage patterns break down as:

• Writing and editing (40%): Editing, communication, document creation
• Practical guidance (24.1%): Tutoring, how-to advice, creative ideation
• Seeking information (13.5%): Factual queries, recipes, current events

That “seeking information” category at 13.5% is the one that directly competes with search engines. Applied to 900 million weekly ChatGPT users, that’s roughly 121 million people using ChatGPT as an information search tool every week. Google handles about 8.5 billion searches daily, so ChatGPT’s search function isn’t replacing Google yet. But it’s carving out a specific niche: the considered, complex question that would have taken 5 to 10 Google searches to answer.

An interesting data point from NBER research: the share of technical help queries in ChatGPT dropped from 18% to 10% between July 2024 and July 2025. Meanwhile, multimedia queries grew from 2% to 7%. People are using ChatGPT less for code debugging (API integrations handle that now) and more for image generation and creative work.

Demographics: who’s searching with AI

Pew Research Center found that 34% of US adults have used ChatGPT as of June 2025, double the 18% from July 2023. The age breakdown tells you where this is heading:

• Ages 18-29: 58% have used ChatGPT
• Ages 30-49: 41%
• Ages 50-64: 25%
• Ages 65+: 10%

The gender split is shifting too. In January 2024, 37% of ChatGPT users had typically feminine names. By July 2025, that share reached 52%. The early adopter skew is evening out.

83% of people who use AI at home use ChatGPT. At work, it’s 71% (Exploding Topics). Google AI Mode is the second most popular in both settings, mostly because it doesn’t require a separate app. You’re already there.

Revenue impact on traditional search

Google still processes roughly 8.5 billion searches daily. That hasn’t collapsed. But the composition of those searches is changing.

The informational queries that drove content marketing for 15 years are getting answered by AI before anyone clicks a link. Transactional and navigational searches remain largely untouched by AI Overviews. Semrush found that transactional keywords make up less than 3% of all keywords triggering AI Overviews. Navigational keywords account for under 2%.

82% of desktop AI Overviews triggered on keywords with fewer than 1,000 monthly searches, per the same Semrush study. That tells you something: AI Overviews aren’t competing on head terms yet. They’re eating the long tail. The specific, detailed queries where content marketers built their traffic base.

Only 5% of SERPs with an AI Overview also show PPC ads. That’s Semrush’s finding, and it has massive implications. If Google is willing to suppress ads on pages with AI Overviews, they’re betting AI answers improve user retention enough to offset ad revenue. That bet tells you where Google thinks search is going.

AI search and marketing spend

50% of US companies have already implemented generative AI in their marketing initiatives, per Capgemini. That’s above the global average of 37%.

The top marketing use cases for AI, according to the same Capgemini CMO survey:

• Data analysis: 62%
• Image and video generation: 59%
• Personalized customer experience: 60%
• Campaign creation: 56%
• Search engine optimization: 56%

That last one is telling. SEO sits equal with campaign creation as a top AI use case. But most of the AI-powered SEO tools focus on traditional search. Generative engine optimization, the practice of structuring content for AI citation, is still new enough that most tools haven’t caught up. That gap is where the opportunity sits for brands willing to invest in AI search optimization now rather than waiting for the market to standardize.

The bottom line for businesses

If you’re reading this in March 2026, here’s what the data says:

• Nearly 1 billion people use ChatGPT weekly. A meaningful chunk of them are asking about products and services in your industry.
• Click-through rates on informational keywords have dropped by half since 2023. AI is answering questions before people click.
• But the traffic that does come through AI referrals converts 4.4x better. Quality over quantity.
• 88% of companies use AI in some function. 27% of workers use it for work. This isn’t early adoption anymore. This is mainstream.

The brands getting cited by AI now are building advantages that compound. Every citation reinforces the pattern. Every recommendation trains future model behavior. The gap between brands visible in AI search and those invisible gets wider with each model update.

If you haven’t checked whether AI engines mention your brand, start with a free AI visibility audit. The data is usually surprising, and not in a good way. You can also browse our latest research on AI search for deeper reads on specific platforms and strategies.