The Gen Z Chatbot Revolution
Gen Z has crossed the tipping point for AI chatbot adoption. Pew Research Center survey data from February-March 2025 shows that 58% of US adults under 30 have used ChatGPT — up from 43% in 2024 and just 33% in 2023. Among all US adults, the figure reached 34%, roughly double the 2023 level.
But the most fascinating insight isn't the headline number — it's how young people use chatbots differently depending on context. Microsoft's analysis of 37.5 million Copilot conversations reveals sharply distinct patterns by device type, time of day, and day of week. On mobile, health and wellbeing dominate. On desktop, work and technology take over. The same person becomes a fundamentally different chatbot user depending on their screen.
ChatGPT Adoption Over Time: Under-30 vs All Adults (US)
Pew Research time series showing accelerating adoption. Note how the under-30 growth curve is steepening, not flattening — +15pp in 2024-2025 vs +10pp in 2023-2024.
Source: Pew Research Center, US adult surveys (2023-2025)
What Young People Actually Ask Chatbots
OpenAI's published research (based on an NBER working paper analysing ChatGPT conversation data) reveals a consistent "intent pyramid": roughly half of all messages are classified as "Asking" — seeking advice, information, or decision support. The other half splits between "Doing" (task completion) and "Expressing" (creative writing, brainstorming).
Specifically, three-quarters of all chatbot conversations cluster into three primary categories: practical guidance and decision support ("What should I do?", "Compare options"), information seeking ("Explain X", "How does this work?"), and writing/drafting ("Write this email", "Summarise this article"). For Gen Z users, these patterns are amplified — they don't just search for information, they use chatbots as personal advisors.
How Users Interact with AI Chatbots: Intent Distribution
OpenAI/NBER research breaks conversations into three modes. "Asking" dominates, accounting for roughly half of all interactions across demographics.
Interaction Modes
Top Usage Categories
Source: OpenAI published summary; NBER working paper on ChatGPT usage (2025)
Device-Context Behavioral Shifts: The Dual-Persona Effect
Microsoft's analysis of 37.5 million Copilot conversations (January-September 2025) reveals what we call the "Dual-Persona Effect": the same user behaves like two entirely different people depending on their device and time of day.
On mobile, health and wellbeing topics dominate consistently — users turn to chatbots for personal health questions, mental wellness support, and relationship advice. On desktop, work and technology dominate during business hours, with topic rhythms shifting between weekdays and weekends. This has profound implications for chatbot designers: the same product must serve very different emotional contexts.
Mobile vs Desktop: Topic Distribution in AI Chatbot Conversations
Microsoft Copilot data from 37.5M conversations shows dramatically different topic profiles based on device type.
Source: Microsoft Copilot usage preprint, 37.5M de-identified conversations (Jan-Sep 2025)
What This Means for Your Business
Youth-First Design
With 58% under-30 adoption, chatbot UX must prioritise mobile-first, conversational, and advisor-style interactions that Gen Z expects.
Context-Aware Responses
Device detection should influence chatbot tone and capability: empathetic on mobile (health context), efficient on desktop (work context).
Accelerating Curve
Growth is steepening (+15pp in 2024-25 vs +10pp in 2023-24). By 2027, under-30 chatbot usage will likely exceed 70%.
Beyond Q&A
75% of conversations involve advice, writing, or decision support — not simple questions. Chatbots must be designed as advisors, not answer machines.
Frequently Asked Questions
58% of US adults under 30 had used ChatGPT by early 2025, according to Pew Research Center data surveyed in February-March 2025. This nearly doubles the 33% figure from 2023, with the growth rate actually accelerating — adding 15 percentage points in 2024-2025 compared to 10 points in 2023-2024.
Extremely fast and accelerating. Under-30 US adults went from 33% (2023) to 43% (2024) to 58% (2025) — a 76% relative increase in two years. Globally, ChatGPT reached 700 million weekly active users by mid-2025, with OpenAI noting especially rapid growth in low- and middle-income countries.
Microsoft's analysis of 37.5 million Copilot conversations shows dramatic device-context shifts: mobile usage is dominated by personal topics (health, wellbeing, relationships), while desktop usage centers on work and technology during business hours. Patterns also shift between weekdays (work-focused) and weekends (personal-focused), suggesting users have dual chatbot personas.
Three categories account for roughly 75% of all conversations: 1) Practical guidance and decision support ("What should I do?", "Compare options"), 2) Information seeking ("Explain X"), and 3) Writing and content creation ("Draft this email"). OpenAI classifies about 50% of messages as "Asking" — the dominant interaction mode across all demographics.
No — it's doing the opposite. The year-over-year increase among under-30s grew from +10 percentage points (2023-2024) to +15 percentage points (2024-2025). With mainstream awareness now achieved and free/open alternatives proliferating, the trajectory suggests 70%+ adoption in this demographic by 2027.
Complete Chatbot Research Series
Global Chatbot Adoption Trends 2026
16.3% global adoption with 3x regional variation and enterprise breakdown.
The AI Chatbot Trust Gap
66% use AI, only 46% trust it — the 20-point deficit explained.
Enterprise Chatbot Barriers
Top 10 pain points: cost (48%), complexity, governance gaps.
AI Chatbot Consumer Scale Wars
Meta AI 1B MAU vs ChatGPT 700M WAU vs Doubao 100M DAU.