AI continues its rapid takeover of tech, media, and culture. It’s in the middle of a classic hype cycle, but that doesn’t mean it’s all hype. It may be oversubscribed, but it’s real and it’s revolutionary. AI’s balanced place in life and business will end up somewhere between zero and its currently-hyped state.
In the meantime, its real traction is evident all around us. That’s mostly something we all see and feel – a tangible but unscientific and anecdotal read. So the question is how we can get a more quantifiable read on AI’s traction. How many people are using it, and how does it stack up against benchmarks like search?
Speaking of search, we continue to track how AI is impacting the previously-dominant front door to the public web. The short version is that AI engines and summaries (e.g., Google AI Overviews) are driving some traffic to online publishers… but not enough to counteract the zero-click searches that AI brings.
But to tell that story requires, again, a quantifiable set of metrics on how much AI is being used. That brings us back to our core question today: How much is AI being used and by whom? To answer that question, we’ve been collecting data points. To save you time, we’ll highlight the best ones here today.
Critical Mass
Our data collection around AI usage and traction recently reached the critical mass required for a data roundup article. So here are the latest stats in no particular order.
– Google AI Overviews now have 2 billion monthly users, up from 1.5 billion in May 2025.
– Google’s Gemini app has 450 million monthly active users with Q2 requests up 50 percent from Q1.
– Google AI Mode has 100+ million monthly active users
– 9 million+ developers have built software using Gemini
– 70 million+ videos have been produced with Google’s Veo 3 AI model since May.
– Google Vids, a Veo-powered feature for text-to-video AI generation now has nearly 1 million monthly active users.
– In Google Meet, 50 million+ people used AI-powered meeting notes
– Across Google products, it processes 980 trillion monthly tokens, up from 480 in May.
OpenAI
– ChatGPT’s has reached 700 million weekly active users, up from 500 million at the end of March.
– Chat GPT has grown in usage 4x year-over-year.
– ChatGPT receives 2.5 billion prompts from global users every day, up from 1 billion in December. This compares with about 14 billion daily searches per day on Google.
AI Apps in General
– In the first half of 2025, users downloaded GenAI apps 1.7 billion times, compared to 1 billion in H2 2024.
– These apps doubled in-app revenue: $1.87 billion in H1 2025, as compared to $932 million in H2 2024.
– Users spend over 15.6 billion hours on GenAI apps in the first half of 2025, a jump from 8.5 billion hours in H2 2024.
– AI platforms in June generated over 1.13 billion referrals to the top 1,000 websites globally – up 357% since June 2024. This still falls far behind Google, which drove 191 billion referrals in the same period.
– 72 percent of teens have tried an AI companion at least once. “Companion” here is defined as AI chatbots designed for personal conversations (e.g., Character.AI), as opposed to utilitarian-focused AI assistants.
– 52 percent of those teens are regular users, and one-third report that the conversations are more satisfying than interactions with human peers.
Absolute Engagement
So there you have it. There’s a lot to unpack in the above stats, but one part that jumps out is search’s dominance. Though AI engines like ChatGPT are growing fast, their absolute engagement is still dwarfed by traditional search. For example, ChatGPT’s 1 billion website referrals versus Google’s 191 million.
That doesn’t mean AI engines don’t pose a real threat to traditional search. Even the small dent so far is meaningful, and it has the potential to grow into something much more disruptive. That will be the part to watch: AI’s engagement relative to search, and how much it deteriorates search volume, and revenue.
We’ll keep a close eye on it.
Header image credit: Luke Chesser on Unsplash
