OpenAI’s ChatGPT has rapidly shifted from an experimental curiosity to a central fixture in the daily lives of millions, according to a new study published by the National Bureau of Economic Research and conducted in collaboration with Harvard economist David Deming. The research, which analyzed more than 1.5 million ChatGPT conversations in a privacy-preserving manner, offers the clearest look yet at how artificial intelligence is being woven into the fabric of society — and not just at work, but at home, in classrooms, and in countless personal moments.
According to OpenAI’s Chief Economist, Aaron Chatterji, these findings reveal that AI is “woven into many aspects of people’s lives,” from answering trivia to offering heartfelt advice. The data shows a striking shift: by June 2025, a staggering 73% of all ChatGPT messages were for personal use, a notable jump from 53% just a year earlier. This surge in non-work usage signals that, for many, AI is no longer just a productivity tool — it’s a daily companion.
The study, published on September 20, 2025, breaks down ChatGPT interactions into three main categories: practical guidance, seeking information, and writing. Practical guidance — think tutoring, how-to help, and creative brainstorming — was the most common use overall. But in the workplace, writing tasks took the lead, accounting for about 40% of work-related messages. Interestingly, most work-related writing wasn’t about creating fresh content from scratch. Instead, users typically asked ChatGPT to edit, critique, or translate existing text. This nuance suggests that, rather than replacing human creativity, ChatGPT is acting more like a supportive editor or a second pair of eyes.
The researchers also examined usage through three distinct lenses: Asking, Doing, and Expressing. Nearly half of all messages fell under “Asking,” showing that many people view ChatGPT as a trusted advisor — a digital confidant, if you will. At work, however, more than half of conversations were about “Doing,” with users employing the chatbot to carry out specific tasks, whether drafting emails, summarizing reports, or troubleshooting code.
Since its launch in 2022, ChatGPT’s potential to disrupt the workforce has been hotly debated. Would it replace jobs, or simply change how we work? The new study leans toward the latter. Researchers argue that ChatGPT is more about enhancing productivity and improving judgment, especially in knowledge-based roles, than about replacing workers outright. Still, as with any sweeping technological change, the impact is complex — and not without controversy.
One area where the study delivers fresh insight is user demographics. As of July 2025, about 10% of the world’s adult population had tried ChatGPT at least once — a figure that underscores the chatbot’s global reach. Early adopters skewed male, but female adoption has since caught up, resulting in a much more balanced user base. Perhaps most strikingly, the fastest growth has occurred in low- and middle-income countries, where adoption rates are four times higher than in wealthier nations. This trend suggests that AI tools like ChatGPT may be leveling the digital playing field, providing access to knowledge and support where it’s needed most.
Yet, as ChatGPT’s influence grows, so do concerns about safety and privacy. In recent months, a string of tragic incidents — including reports of minors committing suicide after forming unhealthy bonds with AI tools like ChatGPT — has heightened scrutiny of generative AI. According to a report published on September 20, 2025, by noted AI safety advocate Eliezer Yudkowsky, these cases have fueled a broader debate about the risks of AI systems forming emotional connections with vulnerable users. Parents, educators, and policymakers are now grappling with how to ensure that AI remains a force for good — especially for young people.
Generative AI has come a long way since its early days, when so-called “hallucination” episodes (in which chatbots invented facts or misled users) were common. Today’s AI bots can generate stunningly realistic images and videos, making it harder than ever to distinguish between what’s real and what’s not. This newfound sophistication brings both promise and peril. On one hand, it unlocks creative possibilities and new ways to communicate; on the other, it raises thorny questions about misinformation, manipulation, and the blurring of reality.
Yudkowsky’s critique of OpenAI’s safety efforts, published in September 2025, argues that the company and its peers must do more to prevent harm — particularly to minors. “We need stronger safeguards to ensure these systems don’t form unhealthy attachments or encourage risky behavior,” he wrote. The call for action echoes across the AI community, where voices on all sides agree that the stakes are rising as generative AI becomes more lifelike and accessible.
OpenAI, for its part, has acknowledged the challenges. The company says it is committed to improving transparency, privacy, and safety, even as it faces stiff competition from rivals like Google’s Gemini and Elon Musk’s XAI. Still, the company’s own research makes clear that ChatGPT is now more than a novelty or a workplace helper — it’s a daily companion for millions, offering advice, comfort, and creativity at the tap of a keyboard.
For many users, ChatGPT is a trusted tutor, a brainstorming buddy, or a late-night sounding board. For others, it’s a tool for translation, editing, or simply getting things done. As the world adapts to this new reality, the line between human and machine advice grows ever blurrier. Is this a revolution in how we learn, work, and connect — or a cautionary tale about the risks of digital intimacy? The answer, it seems, depends on how society chooses to harness the promise and confront the perils of AI.
With adoption soaring in low- and middle-income countries, and a user base that now mirrors the world’s diversity, ChatGPT’s impact is only set to grow. Whether it remains a force for empowerment or becomes a source of new challenges will depend on the choices made by developers, regulators, and users alike. One thing is certain: artificial intelligence, once the stuff of science fiction, is now a daily reality — and the world is watching closely as the story unfolds.
