ChatGPT vs. Claude: How People Actually Use Them & Key Differences
▼ Summary
– Most ChatGPT use is non-work related, with practical guidance, information seeking, and writing tasks accounting for nearly 80% of conversations.
– Claude usage is increasingly automated, with 49.1% of interactions involving minimal user input and 44% of API traffic related to coding tasks.
– AI adoption shows demographic shifts, with ChatGPT’s user base becoming more gender-balanced and skewing younger, while educated professionals use it more for work.
– Geographic usage patterns reveal Israel leads in per-capita Claude usage, while the US shows regional variations with Washington DC having the highest per-capita usage.
– Uneven AI adoption favors wealthier regions, with a 1% higher GDP per capita correlating to 0.7% higher AI usage, potentially widening global inequality gaps.
New research from leading AI developers provides a clear picture of how millions of people are integrating chatbots into their daily routines. While ChatGPT and Claude are both powerful tools, they serve distinctly different purposes for users around the world.
A recent study from OpenAI examined millions of conversations held with ChatGPT between May 2024 and June 2025. The findings reveal a significant shift: non-work usage has now surpassed work-related interactions. What began with nearly half of all messages being job-focused dropped to just over a quarter, even as daily message volume skyrocketed from 451 million to 2.6 billion.
People are turning to ChatGPT primarily for practical guidance, information seeking, and writing assistance. These three categories make up almost 80% of all conversations. Coding, by contrast, represents only a small fraction of usage. When broken down by interaction type, nearly half of all messages involve asking questions, while 40% focus on completing tasks. At work, more than half of ChatGPT use falls into the “doing” category, with over 40% of that dedicated specifically to writing, most often editing text the user already wrote rather than generating entirely new content.
On the other side, Anthropic’s latest Economic Index report shows that Claude is heavily favored for automation, particularly in coding and technical tasks. The study divides usage into two patterns: automation, where the AI completes work with minimal input, and augmentation, where humans and AI collaborate more closely. For the first time, automation has become more common than augmentation, accounting for nearly half of all interactions.
This trend is even more pronounced in business environments. Among enterprise API customers, 77% of conversations follow automation patterns, with the vast majority being directive, users tell Claude what to do and trust the output. About 44% of API traffic relates to coding and mathematical tasks, compared to 36% on the consumer-facing Claude.ai platform, which sees more use in education and writing.
Demographic insights show that ChatGPT now boasts 700 million weekly users worldwide, with growth accelerating in low- and middle-income countries. Gender representation has balanced over time, with women now making up a larger share of active users. Nearly half of all adult messages come from people under 26, and highly educated, well-paid professionals are significantly more likely to use ChatGPT for work.
Claude’s user base tells a different geographic story. The United States accounts for the highest share of usage, but Israel leads in per-capita adoption, with residents using Claude seven times more than expected based on population size. Within the U.S., California dominates in total usage, largely for IT tasks, but Washington, D.C. leads in per-capita usage, mostly for document editing, information retrieval, and job applications.
A concerning pattern emerges when examining economic disparities. Richer countries show higher AI adoption rates, with a 1% higher GDP per capita correlating to a 0.7% higher AI usage score. Wealthier regions also tend to use AI for a broader range of tasks, while lower-income countries concentrate heavily on coding. This has led to warnings about a potential “AI divide,” mirroring historical gaps seen with earlier technologies like electricity and the internet.
In summary, these reports highlight two contrasting approaches to AI interaction. ChatGPT users often treat the tool as a collaborative assistant, asking questions, seeking guidance, and refining their own writing. Claude users, especially in enterprise settings, lean toward automation, issuing directives for coding and technical jobs with minimal back-and-forth. Together, they illustrate not only how AI is being used today, but also how its benefits, and challenges, are being distributed across the globe.
(Source: ZDNET)
