When was the last time you asked ChatGPT for help: drafting an email, looking up a fact, or brainstorming an idea? If you’re like hundreds of millions of others, you’ve probably used it this week. A new analysis from OpenAI and the National Bureau of Economic Research shows that by July 2025, over 700 million people used ChatGPT weekly, sending ~2.5 billion messages per day, roughly 1 in 10 adults worldwide tapping AI for everyday tasks [1][2].

This isn’t a novelty curve; it’s a structural shift. Generative AI has crossed the line from “interesting tool” to daily utility for work and life [1][2].

Global adoption: scale and spread

ChatGPT usage grew ~5× between July 2024 and July 2025 (Diagram 1), reflecting one of the fastest global consumer technology adoptions on record [1][2].

Diagram 1: ChatGPT use exploded between mid-2024 and mid-2025, growing fivefold to 2.5 billion daily messages — showing just how fast generative AI has become part of everyday life.

Table 1: Global growth data (source: [1][2])

But the most striking finding is where the growth is happening. Adoption has been fastest in low- and middle-income countries. In fact, penetration in the lowest-income group of countries jumped from ~10% in 2024 to ~40% in 2025, while high-income countries grew from 25% to 35% over the same period (Diagram 2) [2].

Diagram 2: Adoption grew fastest in low-income countries, where penetration quadrupled, highlighting ChatGPT’s role as a global equalizer in access to knowledge.

Table 2: Adoption by country income (source: [2])

This isn’t just a rich-world phenomenon. For many people in emerging markets, ChatGPT is acting as a first-line knowledge resource, like an all-in-one tutor, translator, and search engine that works in natural language. In places where access to quality education, healthcare information, or professional services can be limited, having a conversational AI that provides instant, low-cost answers helps close information gaps that have existed for decades.

For organizations, this matters because it signals a global customer base that is growing fastest outside traditional tech centers. Companies expanding into emerging markets may find that consumers already use AI as their primary interface with the digital world. For workforce development, it suggests that employees in lower-resource environments are gaining access to the same decision-support tools as their peers in higher-income countries, potentially narrowing skill gaps and creating more level playing fields.

Put simply: adoption in emerging markets shows ChatGPT is not just reinforcing existing advantages in wealthy economies, it may be helping rebalance them [2].

Demographic shifts: from niche to mainstream

The surest sign a technology has crossed into the mainstream isn’t raw scale but who uses it.

In the early months after launch, ChatGPT’s active user base skewed heavily male, roughly 80% of weekly active users had typically masculine first names [2]. By mid-2025, that gap had closed to near-parity, with some months even showing a female majority [2]. This rapid shift from a male-skewed early adopter profile to a balanced user population mirrors the adoption arc of other general-purpose technologies as they move from novelty to everyday utility [2].

Parity matters because it signals normalization: the product is no longer confined to a narrow demographic or specialized use case. A broader, more representative base shapes how the tool is used, what questions are asked, and which features are valued, ultimately influencing product development, content quality, and the expectations people bring to digital experiences [2]. As usage levels out across genders, conversational patterns diversify, prompting styles expand, and demand rises for guidance, safety, and accessibility that work well for everyone, not just early adopters [2].

Diagram 7: ChatGPT adoption moved from heavily male-skewed in early 2024 to near gender parity by mid-2025, showing that the technology has gone truly mainstream. (Source: [2])

This demographic rebalancing underscores a broader truth: ChatGPT is no longer confined to a narrow audience of early adopters. Instead, it has become part of everyday life across different identities, regions, and contexts [2]. With a user base that looks more representative of the global population, the way people interact with the tool has also diversified, shaping not only who uses ChatGPT, but how they use it.

How people use ChatGPT: topics and intents

Across all users, three categories dominate: Practical Guidance, Seeking Information, and Writing. Together, they represent ~77% of usage, with “Seeking Information” rising and “Writing” shifting from creation to editing (Diagram 3) [2].

Diagram 3: Everyday uses dominate: by mid-2025, three topics (i.e., Practical Guidance, Seeking Information, and Writing) made up more than 75% of all conversations.

Table 3: Conversation topics (all users) (source: [2])

This mix of activity highlights a shift in how people approach everyday problem-solving. Practical Guidance requests, like asking for tutoring, brainstorming, or “how-to” support, held steady at ~29%. Seeking Information grew dramatically, from 14% in mid-2024 to 24% in mid-2025, showing that people are increasingly treating ChatGPT like a search engine for complex or personalized questions [2]. Meanwhile, Writing dropped from 36% to 24%, but that doesn’t mean people abandoned it. Instead, writing shifted toward editing and improving user-provided drafts, making ChatGPT less of a creator from scratch and more of a collaborator [2].

At work, the picture sharpens. Writing accounts for 40% of all work-related queries, far ahead of other categories (Diagram 4, Table 4) [2].

Diagram 4: Writing is the #1 workplace use for ChatGPT, accounting for 40% of work queries, followed by guidance and fact-finding. (Source: [2])

Table 4: Work-related topics (source: [2])

This suggests that in professional contexts, ChatGPT is primarily valued as a communication tool: drafting emails, polishing documents, or creating client-facing materials. Practical Guidance (24%) and Seeking Information (14%) are the next most common categories, pointing to its role as a knowledge assistant. Meanwhile, Technical Help, including coding, accounts for 10%, showing that while developers use ChatGPT, the majority of workplace utility lies in text-based and cognitive tasks [2].

Intent is shifting, too. By July 2024, ChatGPT use was evenly split between “Asking” (45%) and “Doing” (45%), with “Expressing” at just 10%. But by June 2025, “Asking” rose to 52%, “Doing” fell to ~35%, and “Expressing” grew to ~14% (Diagram 5, Table 5) [2].

Diagram 5: By mid-2025, “Asking” for advice or information overtook “Doing” tasks, signaling that people now see ChatGPT more as an advisor than just a productivity tool. (source: [2])

Table 5: User intent breakdown (source: [2])

This change reflects a deepening trust in ChatGPT as an advisor, not just an automator. People are turning to it for judgment, recommendations, and interpretation, not only for drafting text or generating outputs. That transition elevates ChatGPT’s role from a productivity tool to a decision-support system, shaping both everyday personal choices and high-stakes professional outcomes [2].

Impact on workplace productivity

Mapping conversations to O*NET activities show ChatGPT’s sweet spot: information work. The top three activities: Getting Information (19.3%), Interpreting Information (13.1%), and Documenting/Recording (12.8%), make up ~45% of work-related usage (Diagram 6, Table 6) [2].

Diagram 6: Nearly half of work-related use is tied to information management, getting, interpreting, and documenting knowledge, the core of modern knowledge work.

Table 6: Top work activities (O*NET) (source: [2])

What does this mean in practice?

• Getting Information (19.3%) reflects tasks like fact-finding, data lookup, or clarifying technical details, functions that once ate up hours of search or reference work now completed in seconds.

• Interpreting Information (13.1%) involves explaining meaning or significance to others, such as summarizing complex reports or turning raw data into client-ready insights. ChatGPT becomes a translator of complexity into clarity.

• Documenting/Recording (12.8%) covers meeting notes, drafting records, or capturing process steps. This helps reduce administrative burdens while improving consistency in knowledge capture.

The next tier of activities, Consulting/Advising (9.2%), Thinking Creatively (9.1%), and Decision-Making/Solving Problems (8.5%), show ChatGPT’s role in more judgment-oriented work. People use it to weigh options, brainstorm solutions, or structure recommendations, reflecting its growing role as a cognitive partner [2].

This is more than task automation, it’s decision-support infrastructure. By accelerating research, synthesis, and documentation, ChatGPT lifts both throughput and quality in knowledge-intensive roles. And because these gains scale with the importance of decisions, the benefits skew higher among highly educated, higher-income professionals, where leverage is greatest [2]. In short, ChatGPT is becoming the scaffolding for how modern knowledge work gets done.

Consumer value and personal empowerment

Over 70% of messages are non-work and that share has been rising, as people use ChatGPT for workout plans, recipes, travel ideas, product research, and creative writing [2]. In the U.S., users would need $98 to forgo ChatGPT for a month. This implies ~$97B in annual consumer surplus in 2024 [2].

For individuals, ChatGPT acts as a daily companion for planning and problem-solving. For organizations, it signals new touchpoints where customers already expect personalized, conversational support [2].

Why this matters

ChatGPT is becoming a daily advisor, helping you learn faster, make decisions, and communicate clearly. Understanding how others use it can help you apply it more intentionally [1][2]. The evidence shows AI is already embedded in workflows, especially in knowledge work. Leaders can capture value by enabling responsible use, codifying best practices, and aligning policy with decision-support use cases that move the needle [2].

As AI becomes a co-pilot for work and life, a crucial question remains: will reliance on AI erode critical thinking, or free people to practice it more? The answer depends on how we implement it, pairing AI speed with human judgment and domain expertise [1][2].