AI Search Engines Take Measurable Share of Informational Queries
Survey data and traffic reports published in mid-2025 confirmed what anecdotal signals had suggested for months: a meaningful share of informational searches, particularly among 18 to 35 year olds and technical professionals, is now going to AI-native search engines rather than Google.
The picture
Perplexity AI had approximately 30 million monthly active users as of April 2025, up from around 15 million at the start of 2024. ChatGPT Search, launched in late 2024, reported similar growth trajectories. Microsoft Copilot, embedded directly into Windows and Edge, added a third significant AI search surface with substantial passive reach.
None of these platforms approaches Google’s scale. Google processes an estimated 14 billion searches per day; Perplexity’s reported query volume remains orders of magnitude smaller. The significance is not current scale, it is the rate of adoption in specific segments and for specific query types.
What queries are shifting
The queries most likely to go to AI-native search tools share a profile:
- Multi-step or research queries. “Help me understand X and then explain how it connects to Y” is better served by a conversational interface than a list of links.
- Code and technical questions. Developers have largely moved this category to ChatGPT and Copilot already; the shift predates ChatGPT Search.
- Comparison and evaluation. “Which tool should I use for X given Y constraints” is a natural fit for AI synthesis.
- Exploratory research. Users who want to build understanding of a new topic, rather than find a specific answer, find conversational interfaces more efficient.
Navigational queries (looking for a specific site), transactional queries (buying something), and local queries (finding a nearby business) remain strongly Google-dominant.
What this means for SEO
The near-term implications for most publishers are modest. Google’s overall query volume continues to grow, and the segments defecting to AI search are not the segments most publishers depend on for traffic.
The medium-term implications are more significant. The under-35 cohort that is most actively adopting AI-native search is also the cohort that will form the majority of the search audience within a decade. Habits formed now will shape the market structure of search in 2030 and beyond.
For SEO strategy today, the most useful response is to ensure content is optimised for retrieval by AI systems, and to build direct audience relationships (email, community, social) that are less dependent on any single discovery platform. The risk is not that Google disappears, but that diversification becomes necessary in ways it wasn’t before.
A note on measurement
Query share data for AI search engines is difficult to verify independently. Most figures come from self-reported platform metrics or panel-based survey data with small sample sizes. The directional signal — meaningful growth in AI search among specific audiences — is credible. Precise market share figures should be treated with caution.