OpenAI's Groundbreaking Study Reveals Shifting Trends in ChatGPT Usage

What do people actually use ChatGPT for? OpenAI provides some numbers.

As someone who writes about the AI industry relatively frequently for this site, there is one question that I find myself constantly asking and being asked in turn, in some form or another: What do you actually use large language models for? Today, OpenAI's Economic Research Team went a long way toward answering that question, on a population level, releasing a first-of-its-kind National Bureau of Economic Research working paper (in association with Harvard economist David Denning) detailing how people end up using ChatGPT across time and tasks. While other research has sought to estimate this kind of usage data using self-reported surveys, this is the first such paper with direct access to OpenAI's internal user data. As such, it gives us an unprecedented direct window into reliable usage stats for what is still the most popular application of LLMs by far. After digging through the dense 65-page paper, here are seven of the most interesting and/or surprising things we discovered about how people are using OpenAI today. OpenAI is still growing at a rapid clip We've known for a while that ChatGPT was popular, but this paper gives a direct look at just how big the LLM has been getting in recent months. Just measuring weekly active users on ChatGPT's consumer plans (i.e. Free, Plus, and Pro tiers), ChatGPT passed 100 million users in early 2024, climbed past 400 million users early this year, and currently can boast over 700 million users, or "nearly 10% of the world's adult population," according to the company. OpenAI admits its measurements might be slightly off thanks to double-counting some logged-out users across multiple individual devices, as well as some logged-in users who maintain multiple accounts with different email addresses. And other reporting suggests only a small minority of those users are paying for the privilege of using ChatGPT just yet. Still, the vast number of people who are at least curious about trying OpenAI's LLM appears to still be on the steep upward part of its growth curve. All those new users are also leading to significant increases in just how many messages OpenAI processes daily, which has gone up from about 451 million in June 2024 to over 2.6 billion in June 2025 (averaged over a week near the end of the month). To give that number some context, Google announced in March that it averages 14 billion searches per day, and that's after decades as the undisputed leader in Internet search. ... but usage growth is plateauing among long-term users In addition to measuring overall user and usage growth, OpenAI's paper also breaks down total usage based on when its logged-in users first signed up for an account. These charts show just how much of ChatGPT's recent growth is reliant on new user acquisition, rather than older users increasing their daily usage. In terms of average daily message volume per individual long-term user, ChatGPT seems to have seen two distinct and sharp growth periods. The first runs roughly from September through December of 2024, coinciding with the launch of the o1-preview and o1-mini models. Average per-user messaging on ChatGPT then largely plateaued until April, when the launch of the o3 and o4-mini models caused another significant usage increase through June. Since June, though, per-user message rates for established ChatGPT users (those who signed up in the first quarter of 2025 or before) have been remarkably flat for three full months. The growth in overall usage during that last quarter has been entirely driven by newer users who have signed up since April, many of whom are still getting their feet wet with the LLM. We'll see if the recent tumultuous launch of the GPT-5 model leads to another significant increase in per-user message volume averages in the coming months. If it doesn't, then we may be seeing at least a temporary ceiling on how much use established ChatGPT users get out of the service in an average day. ChatGPT users are younger and were more male than the general population While young people are generally more likely to embrace new technology, it's striking just how much of ChatGPT's user base is made up of our youngest demographic cohort. A full 46 percent of users who revealed their age in OpenAI's study sample were between the ages of 18 to 25. Add in the doubtless significant number of people under 18 using ChatGPT (who weren't included in the sample at all), and a decent majority of OpenAI's users probably aren't old enough to remember the 20th century firsthand. OpenAI also estimated the likely gender split among a large sample of ChatGPT users by using Social Security data and the World Gender Name Registry's list of strongly masculine or feminine first names. When ChatGPT launched in late 2022, this analysis found roughly 80 percent of weekly active ChatGPT users were likely male. In late 2025, that ratio has flipped to a slight (52.4 percent) majority for likely female users. People are using it for more than work Despite all the talk about LLMs potentially revolutionizing the workplace, a significant majority of all ChatGPT use has nothing to do with business productivity, according to OpenAI. Non-work tasks (as identified by an LLM-based classifier) grew from about 53 percent of all ChatGPT messages in June of 2024 to 72.2 percent as of June 2025, according to the study. Some of this might have to do with the exclusion of users in the Business, Enterprise, and Education subscription tiers from the data set. Still, the recent rise in non-work uses suggests that a lot of the newest ChatGPT users are doing so more for personal than for productivity reasons. ChatGPT users need help with their writing It's not that surprising that a lot of people use a large language model to help them with generating written words. But it's still striking the extent to which writing help is a major use of ChatGPT. Across 1.1 million conversations dating from May 2024 to June 2025, a full 28 percent dealt with writing assistance in some form or another, OpenAI said. That rises to a whopping 42 percent for the subset of conversations tagged as work-related (by far the most popular work-related task), and a majority 52 percent of all work-related conversations from users with "management and business occupations." OpenAI is quick to point out, though, that many of these users aren't just relying on ChatGPT to generate emails or messages from whole cloth. The percent of all conversations studied involves users asking the LLM to "edit or critique" text, at 10.6 percent, vs. just 8 percent that deal with generating "personal writing or communication" from a prompt. Another 4.5 percent of all conversations deal with translating existing text to a new language, versus just 1.4 percent dealing with "writing fiction." More people are using ChatGPT as an informational search engine In June of 2024, about 14 percent of all ChatGPT conversations were tagged as relating to "seeking information." By June of 2025, that number had risen to 24.4 percent, slightly edging out writing-based prompts in the sample (which had fallen from roughly 35 percent of the 2024 sample). While recent GPT models seem to have gotten better about citing relevant sources to back up their information, OpenAI is no closer to solving the widespread confabulation problem that makes LLMs a dodgy tool for retrieving facts. Luckily, fewer people seem interested in using ChatGPT to seek information at work; that use case makes up just 13.5 percent of work-related ChatGPT conversations, well below the 40 percent that are writing related. A large number of workers are using ChatGPT to make decisions Getting help editing an email is one thing, but asking ChatGPT to help you make a business decision is another altogether. Across work-related conversations, OpenAI says a significant 14.9 percent dealt with "making decisions and solving problems." That's second only to "documenting and recording information" for work-related ChatGPT conversations among the dozens of "generalized work activity" categories classified by O*NET. This was true across all the different occupation types OpenAI looked at, which the company suggests means people are "using ChatGPT as an advisor or research assistant, not just a technology that performs job tasks directly." And the rest... Some other highly touted use cases for ChatGPT that represented a surprisingly small portion of the sampled conversations across OpenAI's study: * Multimedia (e.g. creating or retrieving an image): 6 percent * Computer programming: 4.2 percent (though some of this use might be outsourced to the API) * Creative ideation: 3.9 percent * Mathematical calculation: 3 percent * Relationships and personal reflection: 1.9 percent

Most People Use ChatGPT for Personal Life, Not Work, According to a New OpenAI Study

Macy has been working for CNET for coming on 2 years. Prior to CNET, Macy received a North Carolina College Media Association award in sports writing. When ChatGPT first launched in November 2022, parent company OpenAI pitched it as a productivity tool and a game-changer for delegating menial work tasks, such as responding to emails or writing memos. OpenAI just released a new paper looking into how hundreds of millions of people globally actually use ChatGPT, and the results show a striking shift in how people use it. What started as a work assistant is now a tool people use for their personal lives. In mid-2024, nearly half of all conversations on ChatGPT were job-related. By mid-2025, that number had fallen to just over a quarter. Don't miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source. That doesn't mean people are using the chatbot less. ChatGPT has about 700 million weekly active users worldwide, sending more than 2.5 billion messages per day or about 29,000 messages per second, according to the report by the National Bureau of Economic Research and contributors from OpenAI, Duke University and Harvard University. In short, more and more people are using the platform -- just not for work-related queries. Read also: OpenAI Wants You to Get a Certificate in ChatGPT and Find Your Next Job The paper describes a shift away from asking ChatGPT to perform tasks, such as writing text, toward users asking it questions. For instance, "writing help comprised more than a third of usage last year. Now, it's closer to a quarter." Meanwhile, "seeking information" has grown from 14% to 24% of all conversations, meaning people use ChatGPT more as a search engine replacement for information and guidance. "Overall, our findings suggest that ChatGPT has a broad-based impact on the global economy," the paper states. "The fact that non-work usage is increasing faster suggests that the welfare gains from generative AI usage could be substantial." (Disclosure: Ziff Davis, CNET's parent company, in April filed a lawsuit against ChatGPT maker OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.) The OpenAI paper also looks into the demographics behind these inquiries:

OpenAI: Women Make Up 'More Than Half' of ChatGPT Users

Don't miss out on our latest stories. Add PCMag as a preferred source on Google. OpenAI is pulling the curtain back on what people are talking to ChatGPT about in a new research paper that acknowledges "little is known" about how large language model (LLM) chatbots are being used. The 63-page report contains many interesting nuggets, but two points are particularly surprising. First, although most early adopters were men, "more than half of weekly active users" are women as of June 2025, classified by researchers as those with "typically female first names." Still, usage is growing rapidly across genders and even countries. OpenAI claims 700 million users -- 10% of the global adult population -- are now using ChatGPT weekly. Second, 73% of all chats are not work-related, a sharp increase from 53% a year ago. Personal and professional chats "have grown continuously, but non-work messages have grown faster." The most common conversation topics -- "practical guidance," "seeking Information," and "writing" -- account for 78% of all conversations." Writing is number one when it comes to work-related tasks. Vibe coding, or asking an AI to generate computer code, represents a somewhat small share of ChatGPT messages at 4.2%. It's possible programmers prefer other tools, such as Claude Code or coding-specific AI agents like Replit. Self-expression is also a surprisingly low use case, defined as "greetings and chitchat, relationships and personal reflection, [and] games and roleplay." That's good news given the risks we've seen in people developing personal relationships with AI chatbots. The data comes from a random selection of messages sent to ChatGPT on its consumer plans (not enterprise or education), including ChatGPT Free, Plus, and Pro, between May 2024 and June 2025. The paper's authors say they did not personally read any user messages, but 46% were from those ages 18-25. The paper's goal is to clarify the chatbot's economic impact. The first co-authors listed are Aaron Chatterji, the chief economist at OpenAI and a professor at Duke University, Tom Cunningham, an economic researcher at OpenAI, and David Deming, a political economist from Harvard. Given the findings that people are not using ChatGPT for their jobs, the paper seeks to find other economic value for the tech. "Overall, we find that ChatGPT provides economic value through decision support," the report says, later adding: "Our findings suggest that ChatGPT has a broad-based impact on the global economy. The fact that non-work usage is increasing faster suggests that the welfare gains from generative AI usage could be substantial."

OpenAI releases first-of-kind study revealing how people are using ChatGPT for everyday tasks

Illustration of the ChatGPT App on the iOS App Store displayed on a phone screen. Despite rapid adoption of large language models like OpenAI's ChatGPT, few comprehensive studies have delved into exactly how the technology is being used in everyday life -- that is, until now. On Tuesday, researchers, including those from OpenAI, released a first-of-its-kind study that examines who was using ChatGPT and for what purposes based on internal messages sent to ChatGPT on consumer plans. Amongst the key findings from the National Bureau of Economic Research working paper was the surge in non-work related messages, jumping to 73% in June 2025 from 53% a year ago. In a post on LinkedIn, OpenAI Chief Economist Aaron Chatterji said that the shift was a signal that ChatGPT is becoming part of many aspects of people's lives. "We're still learning how people use AI in the wild, but this trend gives us a glimpse into where the value is and how it's shifting," he added. The study, which has not been formally scrutinized by other researchers, or peer reviewed, was authored by OpenAI's Economic Research team and Harvard economist David Deming, and said to have drawn on "a large-scale, privacy-preserving analysis of 1.5 million conversations." Here is an outline of some of the most notable revelations:

OpenAI reveals how people really use ChatGPT: personal tasks over work, women users outnumber men

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. In brief: There are around 700 million people actively using ChatGPT each week, sending around 29,000 messages per second - so what are these people actually using the chatbot for? OpenAI has just released an extensive report that looks into the question. The biggest takeaway is that ChatGPT is now primarily used for non-work purposes, and female users appear to outnumber male users. The National Bureau of Economic Research (NBER) working paper by OpenAI's research team and Harvard economist David Deming analyzed 1.5 million conversations to track how ChatGPT's use has evolved in the three years since its launch. Like all generative AIs, ChatGPT arrived with the promise of revolutionizing the way we work. In June 2024, 53% of usage by non-enterprise accounts was non-work related; in June 2025, that figure had risen to 73%, with conversations focused on practical guidance, seeking information, and writing. Most people use the chatbot for "Practical Guidance." This category, which made up 28.8% of the sampled conversations, encompasses tutoring or teaching, "how to" advice, health, fitness, beauty, or self-care, and creative ideation. The second most popular category was Seeking Information, which has grown from 14% to 24%. Most of the section is made up of people trying to find specific info. There are also those looking for purchasable products and cooking/recipes. The findings highlight how more people are choosing ChatGPT as an internet search engine over traditional methods like Google. Writing had been the most popular use for ChatGPT in July 2024, but it has since fallen to third. The largest use case in this section was for editing or critiquing provided text (10.6%), more than the 8% who use it for personal writing or communication such as emails. A lot of people use it for translation purposes, argument or summary generation, and fiction writing. When it comes to how ChatGPT is utilized in the work environment, most people use it for documenting and recording information. This is followed by making decisions and solving problems, thinking creatively, and working with computers. Getting information is the least common use case at work, which could be due to the risk of hallucinations. OpenAI wrote that the results mean people are "using ChatGPT as an advisor or research assistant, not just a technology that performs job tasks directly." As for the demographics that are using ChatGPT, it appears that female consumer users now outnumber men. In January 2024, only 37% of users had names that are typically considered feminine. In July this year, that figure had increased to 52%. ChatGPT is certainly most popular among younger people, with almost half (46%) of users self-reporting as being between 18 and 25. This group is more likely to ask for advice or about hobbies, while older users opt for more work-related queries. Moreover, people with college degrees or higher, as well as those in highly-paid or technical jobs, are more likely to use the tool for work. While almost 10% of the world's adult population now use ChatGPT, the number of weekly active users on its consumer plans is still growing at a rapid pace. This is increasing the number of messages the AI is processing, from around 451 million in June 2024 to over 2.6 billion in June 2025. However, much of the recent growth is coming from new signups. Based on per-user message rates, long-term account holders often go through lengthy periods where their number of daily messages are unchanged, only increasing following the launch of a new feature or upgrade.

OpenAI Reveals How (and Which) People Are Using ChatGPT

Large language models largely remain black boxes in terms of what is happening inside them to produce the outputs that they do. They have also been a bit of a black box in terms of who is using them and what they are doing with them. OpenAI, with some help from the National Bureau of Economic Research (NBER), set out to figure out what exactly its growing user base is getting up to with its chatbot. It found a surprising amount of personal use and a closing "gender gap" among its frequent users. In an NBER working paper authored by the OpenAI Economic Research team and Harvard economist David Deming, the researchers found that about 80% of all ChatGPT usage falls under one of three categories: “Practical Guidance,†“Seeking Information,†and “Writing." "Practical guidance," which the study found to be the most common usage, includes things like "tutoring and teaching, how-to advice about a variety of topics, and creative ideation," whereas "seeking information" is viewed as a substitute for traditional search. "Writing" included the automated creation of emails, documents, and other communications, as well as editing and translating text. Writing was also the most common work-related use case, per the study, accounting for 40% of work-related messages in June 2025, compared to just 4.2% of messages related to computer programmingâ€"so it seems coding with ChatGPT is not that common. Notably, work usage for ChatGPT appears to make up a shrinking share of how people are interacting with the chatbot. In June 2024, about 47% of interactions users had with the chatbot were work-related. That has shrunk to just 27%, which comes as other research shows companies largely failing to figure out how to generate any sort of meaningful return from their AI investments. Meanwhile, non-work-related interactions have jumped from 53% to 73%. While users are apparently spending more time with ChatGPT in their personal time, OpenAI's research found that a "fairly small" share of messages with the chatbot were users seeking virtual companionship or talking about social-emotional issues. The company claimed that about 2% of all messages were people using ChatGPT as a therapist or friend, and just 0.4% of people talked to the chatbot about relationships and personal reflectionsâ€"though it'd be interesting to see if users who engage with a chatbot this way generate more messages and if there is stickier engagement. For what it's worth, other researchers seem to believe that this usage is far more common than those numbers might suggest. Common Sense Media, for instance, found that about one in three teens use AI chatbots for social interaction and relationships. Another study found that about half of all adult users have used a chatbot for "psychological support" in the last year. The teen figure is particularly of note, considering OpenAI's research did find its userbase skews young. The NEBR study found 46% of the messages came from users identified as being between the ages of 18 and 25 (it also excluded users under the age of 18). Those users are also more likely to use ChatGPT for personal use, as work-related messages increase with age. The study also found that there is a growing number of women using ChatGPT, which initially had a very male-dominated user base. The company claims that the number of “masculine first name†users has declined from about 80% in 2022 to 48% in June 2025, with "typically feminine names" growing to reach parity. One caveat about the study that may give you pause, depending on how much you trust technology: OpenAI used AI to categorize all of the messages it analyzed. So if you're skeptical, there's an asterisk you can put next to the figures.

Here's what the data says people ask ChatGPT

OpenAI released the first detailed public study on who uses its chatbot and what they most often ask it to do. SAN FRANCISCO -- ChatGPT-maker OpenAI released the first detailed study of what its users do with the popular chatbot and who they are, providing an unprecedented look at how people use the artificial intelligence tool and what they talk to it about. The company reports that most ChatGPT users are women and that the majority of requests sent its way are not work-related. The user base is dominated by young people -- nearly half of the conversations studied were from people aged 18 to 25.

1.1 million ChatGPT messages analyzed -- here's what most people are asking

A new study from OpenAI analyzed more than 1.1 million real ChatGPT messages, and it turns out, the primary use for the AI tool is not at the office. ChatGPT is now being used more for everyday tasks, such as staying fit, preparing meals and helping kids with homework. The researchers looked at how usage evolved between May 2023 and April 2024, a period that saw enormous growth in ChatGPT adoption and discovered that work-related use is way down; usage dropped from 57% of messages to just 28%. That doesn't mean we're using ChatGPT less. We're just using it differently. Instead of firing off prompts to help meet deadlines or decode spreadsheets, people are turning to ChatGPT for advice, planning and even emotional support. The study found that most prompts now fall into three major areas: Guidance. Users are looking for help in their lives, such as how to write a compelling Substack post to engage their audience, how to fix a squeaky door or even how to break up with someone. Information. Users are using ChatGPT for answers that they might have Googled years ago. Questions about their health, the best time to travel or how to know if food has expired. Writing help. While this may sound like something needed at the office, users are turning to ChatGPT to help with their dating profile, apology letters or crafting the perfect party invitation. And while coding might seem like a major use case, it actually makes up less than 5% of all messages. Among the biggest takeaways from the study is this: 1 in 4 ChatGPT messages are now "information-seeking." In other words, we're now turning to ChatGPT instead of Google. The personalized response customized to each user's exact needs is essentially why more people are turning to the chatbot for answers over the search engine, or even tools like Google Lens. The study emphasized that ChatGPT usage is growing fastest in places like Brazil, India and Nigeria, where mobile-first access to answers matters. For a long time, ChatGPT skewed male -- early adoption patterns and all that. But the gender divide is finally closing. According to the study, users with feminine-coded names slightly outnumber male ones now, and the prompts themselves reflect that broader demographic. People are asking about parenting, relationships, meal planning, budgeting and even mental health (although therapy-style chats remain a small portion of overall use). Clearly, ChatGPT has become much more of a daily companion and is no longer confined to the 9-5 workday. Studies like this give us insight into how AI tools are being used while also confirming what we personally already know. ChatGPT is a great source of information, especially when we can't be bothered with sponsored links, ads or excess information. ChatGPT is becoming a daily companion that goes with us beyond our desk and into our kitchens, garage, classroom, calendars and even late-night overthinking spirals. For many of us, this study isn't surprising; it's just finally been measured, proving we aren't alone.

OpenAI reveals biggest-ever study of how people are using ChatGPT - here are 3 things we've learned

The survey was conducted by OpenAI's Economic Research team and Harvard economist David Deming Open AI just revealed the results of its largest-ever ChatGPT usage study to date, and the insights into how people interact with the AI chatbot are fascinating. Revealing the study on its blog, OpenAI broke down the results into three key takeaways: Who's using ChatGPT, what they're using it for, and how usage of the AI is evolving over time. The study was conducted by OpenAI's Economic Research team and Harvard economist David Deming for the National Bureau of Economic Research (NBER). While Future (TechRadar's parent company) has conducted its own research in the past that has given a great insight into the wider population's AI usage, getting information directly from the creators of the world's most popular AI tool gives an excellent insight. Here are three things we learnt from the world's biggest study into ChatGPT and how people use it. According to OpenAI, the gender gap of users taking full advantage of ChatGPT has closed dramatically over the years. Previous research in January 2024 showed of the users who had names that it was possible to classify as masculine or feminine, 37% had "typically feminine names." As of July 2025, however, that figure stands at 52%. Now, I have my doubts as to how OpenAI can truly determine the gender of a user interacting with ChatGPT, but considering the massive influx with the same criteria, it's fair to say that there is far less of a gender gap related to AI than ever before. Elsewhere, OpenAI reports that the adoption rate in the lowest-income countries is 4 times faster than that of the wealthiest. According to OpenAI, the main reason people use ChatGPT is to get things done, and "three-quarters of conversations focus on practical guidance, seeking information, and writing." OpenAI breaks down how people use ChatGPT into three categories: Asking, Doing, and Expressing. The study claims nearly half of all ChatGPT queries are "Asking" - where a user interacts with AI to get advice. 40% of usage is "Doing" such as "drafting text, planning, or programming," and 11% of usage is "Expressing", focused on "personal reflection, exploration, and play." While this breakdown might not surprise you, I think it's interesting to see that writing is in fact the most common work task people use ChatGPT for. Honestly, that doesn't fill me with much hope, especially considering I spend hours and hours of my day writing without help from AI. This part of the study is the most intriguing, in my opinion. In fact, I'm actually pretty shocked at the results. OpenAI claims 70% of people use ChatGPT in a non-work related fashion, with the study stating "ChatGPT helps improve judgment and productivity." Without this study, I'd have honestly believed most people use ChatGPT in a work-based environment, specifically for mundane tasks that don't require a lot of brain power. Instead, it looks users are interacting with ChatGPT after the office closes and in their spare time, highlighting just how much AI has managed to infiltrate our daily lives. You can read the full paper with even more statistics on the NBER website. What result from this survey do you find the most interesting? Let us know in the comments below.

Women use ChatGPT as much as men do

Why it matters: Closing the gender gap helps ensure women aren't left behind as AI reshapes work and life. In the first few weeks after its release, OpenAI estimated that as many as 80% of users were male, according to the report. * More recent data from June found slightly more usage among those with traditionally female names. * OpenAI estimates its gender mix by analyzing user names as typically male, female or uncertain, which can indicate broad trends but isn't a scientifically accurate metric for assessing usage by gender. * The key to gender balance has been expanding usage beyond early adopters, OpenAI chief economist Ronnie Chatterji told Axios. "There's been so much excitement about ChatGPT and how people can use it to do really practical things," he said. By the numbers: OpenAI's economic report draws on a sampling of 1.5 million conversations from ChatGPT's roughly 700 million weekly users. * 80% of usage fell into three categories: practical guidance, information search and writing help. * Coding and other specialized uses were far less common. * Using ChatGPT as a companion or virtual therapist accounted for less than 2% of usage, per the study. The fine print: OpenAI studied a sampling of messages from logged-in users over 18, excluding those who specifically opted out, as well as those who requested their data not be used to train new models.

OpenAI's biggest research on human-ChatGPT talks will surprise you

Think a lot of people are secretly pouring their heart out to an AI chatbot? Nope. OpenAI has released its biggest analysis to date about the usage habits of ChatGPT. In a research paper courtesy of the company's Economic Research team and Harvard economist David Deming, the company analyzed 1.5 million ChatGPT conversations, and it's now shedding light on some interesting stats about how interactions with the AI chatbot unfold. The who, why, and how much? As of July 2025, ChatGPT had over 700 million users exchanging 18 billion messages each week, or 2.5 billion messages per day. And it seems people are using ChatGPT for work-related assistance more than anything else. "Writing is by far the most common work use, accounting for 42% of work-related messages overall and more than half of all messages for users in management and business occupations," says the research paper. Compared to the early launch days, the gender gap has reversed, and more than half of users now have a female name. Additionally, nearly half of all ChatGPT users are under 26 years of age. Practical guidance, writing, and seeking information are three of the biggest discussion areas with ChatGPT, accounting for 78% of all exchanges. In 49% of the messages, users were asking (of messages are users asking ChatGPT for guidance, advice, or information), while 40% accounted for doing (tasks connected into workflow), and A few surprises In the tech circle, it's often said that programmers are at risk of losing their jobs to AI, more so than any other profession. Yet, questions about computer programming and coding only account for 4.2% of the analyzed chats. Over the past few weeks, ChatGPT's negative impact and a few deadly incidents have resulted in hot debates, prompting the company to build parental controls and warning systems for guardians. Yet, Relationships and Personal Reflection messages only accounted for 1.9% of the assessed conversations. In only 1% of the analyzed conversations, users were expressing themselves to the AI chatbot. That's pretty surprising, considering reports of how users can get sucked into romantic conversations with ChatGPT, a pattern that is even more evident with AI personality bots available on platforms such as Nomi, CharacterAI, and Replika.

Here's what's in OpenAI's largest study on ChatGPT users

ChatGPT is the most popular AI chatbot in the world. At the end of July, its weekly active users made up nearly 10% of the world's adult population, a report found. OpenAI's economic research team -- the chatbot's owner -- with a Harvard economist released Monday a National Bureau of Economic Research working paper diving into who is making up those 700 million accounts and how they're using the chatbot. The study analyzed 1.5 million internal conversations with ChatGPT from consumer plans. According to the report, the reviewed messages were not seen by humans and were only reviewed by an LLM. "This suggests that gender gaps in ChatGPT usage have closed substantially over time," the report said. However, the study noted that names not in the datasets used as reference for the LLMs or that were flagged as "ambiguous" were marked as "unknown" and excluded from the report. The study also found that ChatGPT users are skewing younger. Based on users who reported their age, nearly half of the messages reviewed were sent by adults aged 18 to 25. Highly educated users in professional occupations were found to be more likely to use ChatGPT for work. More than 46% of users with a college degree use ChatGPT for work compared to 37% of those without a college degree. The study found that people in highly paid, technical roles are more likely to use ChatGPT for work, especially those in computer-related jobs. ChatGPT is growing quickly in low- and middle-income countries, too. The study found that by May of this year, the chatbot was adopted in the "lowest income countries" at four times the rate of the "highest income countries." When using the chatbot, about 77% of conversation topics revolve around practical guidance, writing, and seeking information. The study defines practical guidance as how-to advice, teaching, creative ideation, and more. Writing includes editing, translation, and summary generation, among others. And seeking information includes asking for specific information and recipes, among others. Users that ask ChatGPT for technical help -- which includes calculations, data analysis, and computer programming -- dropped from 12% in July 2024 to about 5% in July 2025. And when split among gender, the study found that users with "typically female first names" are more likely to use ChatGPT for writing and practical guidance while users with "typically male first names" use the chatbot more for technical help, seeking out information, and for multimedia -- like for creating images. The majority of messages to ChatGPT -- about 49% -- revolve around asking the chatbot for guidance, advice, or information while another 40% focus on asking the chatbot to complete a task. Only 1% have "no clear intent." Among users under the age of 26 -- which make up 46% of the data -- about 23% of them use the chatbot for work. Writing makes up 42% of work-related messages as of July 2025 -- with more than half coming from users in management and business occupations. These messages are primarily asking ChatGPT to modify text instead of creating "text from scratch," the report found.

The most common reasons why people use ChatGPT

In a first-of-its-kind study, ChatGPT's parent company OpenAI says it knows what people are asking its chatbot to do. People mostly use ChatGPT for everyday tasks, according to a new study by the artificial intelligence (AI) chatbot maker OpenAI. A random sample of 1.5 million conversations between users with ChatGPT between May 2024 and June 2025 was analysed to track how people are using it, for the study that the company calls "the most comprehensive study of actual consumer use of AI ever released." It found that the chatbot was used 30 per cent of the time for work-related tasks and 70 per cent for non-work-related tasks. The most common non-work conversations typically focus on "everyday tasks" that create "economic value," in people's personal and professional lives, it said. In three out of four conversations, the focus of the prompt for ChatGPT was on "practical guidance, seeking information and writing," the company wrote. Almost half of all the messages were classified as "asking" ChatGPT a question, which means users "value ChatGPT most as an advisor," rather than just for task completion, the study said. Any prompts that asked ChatGPT to complete a task such as writing text, planning or programming, made up about 40 per cent of the messages. The other 11 per cent were tasks that involved "personal reflection, exploration and play". Just under 2 per cent of all messages were related to companionship or social-emotional issues, the report found. Writing dominates professional tasks in ChatGPT Those who use the chatbot for work-related questions are more often "educated users in highly-paid professional occupations," who use it as an advisor or researcher instead of asking it to do the entire task from start to finish, it added. Writing was the largest share of professional tasks that the chatbot was asked to do, but most of the time it was asked to modify user text either through editing, critiquing or translating instead of generating new text from scratch, the study found. Other work tasks, such as coding and self-expression, are still "niche" activities that the chatbot performs, the company added, with just 4 per cent of all queries to ChatGPT related to computer programming. One of the key benefits of using ChatGPT in a work scenario is supporting a user with their decision-making because the app "improve[s] judgment and productivity, especially in knowledge-intensive jobs," OpenAI said. The report also noted that women are starting to use ChatGPT more than they used to. It said a previously reported gender gap has "narrowed dramatically," with 52 per cent of the chatbot's users having female-sounding names in 2025, up from 37 per cent in 2024. To study the messages, OpenAI used automated tools that categorised the nature of the conversation so humans didn't need to review the content. Harvard economist David Deming teamed up with OpenAI's Economic Research team to analyse the content.

OpenAI wants to transform business. Many of its users just want life hacks

During its early years, OpenAI looked like it might build a business selling access to its increasingly powerful AI models to Fortune 500 companies. But when ChatGPT launched (almost by surprise) in late 2022, the startup suddenly had a breakout consumer product -- one that raced to 100 million users in just a few months, faster than any app in history. Overnight, OpenAI became a consumer tech brand and, most importantly, the poster child for generative AI in the minds of everyday users. Today, ChatGPT has more than 700 million weekly active users worldwide, according to OpenAI. And the way those people use the chatbot suggests the company may be drifting further toward the consumer market. This week, OpenAI released a study of 1.5 million user chat logs between May 2024 and June 2025, revealing that nearly three-quarters (73%) of chats were personal rather than work-related. Just a year earlier, in June 2024, personal and work prompts had been roughly equal. (That data excludes OpenAI's API customers, who are largely developers and enterprises.) The report comes at a time when, across industries, many enterprises are growing skeptical about how -- and when -- AI tools might deliver the efficiencies they were promised, the kind executives can tout on earnings calls. Despite the hype, by most objective accounts, the AI transformation hasn't yet materialized. An August MIT report, for example, found that 95% of enterprise AI pilot projects have stalled. Meanwhile, talk of an AI bubble continues, with critics raising an eyebrow at bullish startup valuations and tech stock prices.

What Has ChatGPT Become?

For the first time since its debut in 2022, a large trove of user data from ChatGPT has been made available to researchers, who attempted to answer a straightforward -- but contentious and heavily obfuscated -- question: What are most people doing with ChatGPT, most of the time? There are a few major caveats here. OpenAI's own researchers worked on the paper with Harvard economist David Deming under the auspices of the National Bureau of Economic Research (in other words, the company is comfortable with this paper's findings). Additionally, the research, which is mostly an attempt to classify and sort a large set of messages, was done substantially by OpenAI's own models, which both automated the process and, the researchers say, helped preserve user anonymity. "No one looked at the content of messages while conducting analysis for this paper," the researchers claim, although they did validate their automations against human analysis used in smaller previous studies we've talked about here before. The researchers' methodology is both fascinating and could insufficiently but not entirely inaccurately be characterized as "we asked ChatGPT." With all that in mind, here are a few of the researchers' findings: as of July, "about 70% of ChatGPT consumer queries were unrelated to work," with non-work queries increasing faster. The three most common topics of conversation are what the company and researchers call Practical Guidance, Writing, and Seeking Information (which the company says "appears to be a very close substitute for web search), with Computer Programming and Relationship and Personal Reflections making up a small percentage of messages. Within the category of work, writing assistance is by far the main use case, with text modification or editing, rather than composing from scratch, making up a majority of requests. Most interesting is the paper's attempt to categorize what "kind" of output users are looking for: About 49% of messages are users asking ChatGPT for guidance, advice, or information (Asking), 40% are requests to complete tasks that can be plugged into a process (Doing), and 1% are messages that have no clear intent (Expressing). Asking messages have grown faster than Doing messages over the last year and are rated higher quality using both a classifier that measures user satisfaction and direct user feedback. None of this will be too surprising to anyone who has been using tools like ChatGPT for more than a few months, or even to people who are merely aware of how their friends, family, and coworkers seem to be chatbotting. (While the paper describes widespread uses that are compatible with "doing your homework for you," its design allows it to generally steer around the topic.) It is, however, slightly out of step with some of the more provocative messaging that companies like OpenAI have at times engaged in or at least indulged. The paper describes a powerful, versatile tool that people use for a wide variety of assistive tasks -- something akin to a search engine with powerful automated productivity software attached to it -- rather than an increasingly autonomous human-like entity. It describes a technology that's become less alien, not more, to its users, and notes their evolving expectations. "Within work usage, we find that users currently appear to derive value from using ChatGPT as an advisor or research assistant, not just a technology that performs job tasks directly," the paper concludes, arguing that it "likely" improves worker output by offering "decision support," or advice. The research brings to mind another paper, published last year in Science, in which an interdisciplinary group of researchers argues that LLMs are better understood as "cultural and social technologies" than in terms of intelligence and autonomy, making comparisons to search, bureaucracy, and even the written word -- systems that share the trait of "allowing humans to take advantage of information other humans have accumulated" in new and transformative ways. The picture that emerges from this data matches this thesis pretty closely: ChatGPT, for many of its users, is a way to access, remix, summarize, retrieve, and sometimes reproduce information and ideas that already exist in the world; in other words, they use this one tool much in the way that they previously engaged with the entire web -- arguably the last great "cultural and social technology" -- and through a similar routine of constant requests, consultations, and diversions. One doesn't get the feeling from this research that we're careening toward uncontrollable superintelligence, or even imminent invasion of the workforce by agentic AI bots, but it does suggest users are more than comfortable replacing and extending many of their current online interactions -- searching, browsing, and consulting with the ideas of others -- with an ingratiating chatbot simulation.

Here's How People Are Actually Using ChatGPT, According to OpenAI

Since its launch in November 2022, ChatGPT has changed the way people write emails, manage their social media accounts, and generate code. Now, a new report from ChatGPT-maker OpenAI is giving fresh insight into how people are really using the chatbot. OpenAI's researchers published a 64-page study on Monday with the National Bureau of Economic Research (NBER) that found nearly 80% of all conversations with ChatGPT were concerned with three categories: practical guidance, seeking information, and writing help. The study, which was based on more than one-and-a-half million ChatGPT messages sent from May 2024 to July 2025 by 130,000 users, is the largest of its kind to date. Here's what it found: Related: ChatGPT's New Update Can Create PowerPoint Presentations and Excel Spreadsheets for You ChatGPT's demographics have changed in the years since its launch. The percentage of male users has declined from 80% in the first few months after the chatbot's drop in late 2022 to 48% as of June 2025, which makes the chatbot's primary user base now primarily female. Meanwhile, nearly half of all messages sent to the chatbot since launch were sent by users under the age of 26. Gen Z is embracing AI, with a survey conducted by International Workplace Group earlier this year finding that close to two-thirds of Gen Z respondents were teaching their older colleagues how to use AI. OpenAI also announced on Tuesday that it was creating a different version of ChatGPT for teen users under the age of 18 that prioritized teen safety. The most common use case was "practical guidance," which is defined in the report as encompassing activities, such as tutoring, teaching, how-to advice, and coming up with creative ideas. The next most popular category involved "seeking information," which is labeled as searching for information about people, products, and current events, like conducting a web search. The final popular use case included writing tasks that automatically generate emails and documents, and editing text. Writing was the most common use case at work, with an average of 40% of work-related messages on ChatGPT stemming from writing queries. Most requests asked ChatGPT to look at text the user had already written instead of creating something new. In other words, two-thirds of writing messages asked ChatGPT to edit, translate, critique, or modify text instead of generating new text. "Writing dominates work-related tasks, highlighting chatbots' unique ability to generate digital outputs compared to traditional search engines," the study read. Related: ChatGPT's Creators Are Worried We Could Get Emotionally Attached to the AI Bot, Changing 'Social Norms' The study also classified messages another way, using three categories based on the kind of output the user was looking for: Asking, Doing, or Expressing. "Asking" applies to nearly half of all messages sent to the chatbot (49%), which occurs when a user seeks information about a subject or a solution to a problem. "Doing" refers to tasks where a user wants an output, especially writing activities, and applies to 40% of all messages. "Expressing," which refers to 11% of all messages, happens when a user communicates their views or feelings without asking for any information or action. Based on these classifications, users were more likely to use ChatGPT to find answers to questions rather than to carry out tasks or express opinions. ChatGPT users tapping into the chatbot at work were most likely to use it to seek information and find information. "Overall, we find that information-seeking and decision support are the most common ChatGPT use cases in most jobs," the study reads. Related: Is Your ChatGPT Session Going On Too Long? The AI Bot Will Now Alert You to Take Breaks Despite discussion from tech leaders like Google CEO Sundar Pichai and Nvidia CEO Jensen Huang, it seems like vibe coding is still niche. The study found that comparatively few users were tapping into ChatGPT to code; only 4.2% of messages were related to computer programming, much less than 33% of all work-related conversations with competing chatbot Claude from Anthropic. Also, only a small percentage of users were using ChatGPT for companionship or guidance on social issues. Less than 2% of ChatGPT messages were about relationships and personal reflection, per the study.

OpenAI Spills The Beans On ChatGPT: How People Worldwide Are Using The Tool For Writing, Learning, And Daily Life

When OpenAI first introduced ChatGPT, the AI chatbot, many were uncertain about its practical use and questioned whether it had any real purpose and value. Despite the initial wondering, the platform gained momentum quickly, and fast forward to today, it has expanded extensively, and users are relying on the platform for more than just simple questions and answers. A recent OpenAI study sheds some light on the diverse ways the tool is being used and how accessible it has become when it comes to learning, creativity, and problem-solving tasks. Despite the initial skepticism around ChatGPT and questions regarding any practical use the AI tool would have, it has now been so widely adopted that many people are relying on it for their daily tasks, and even going deeper than that. OpenAI has now released its first large-scale study, which has data collected over an entire year, and it covers more than a million messages to answer that question. The findings highlight how people around the world have blended the chatbot into their routines and have it serve as more of a practical companion. One of the study's major findings is that writing and editing remain the fundamental tasks users seek ChatGPT for. The tool is not used to create content from scratch, but rather to help refine drafts and make the needed edits. This is vital because apprehensions regarding the tool replacing human creativity have been there for quite some time, and this adds clarity to the air of confusion by highlighting how the tool is relied on more for adjusting tone, structural improvements, and making the user feel more confident about their final work. ChatGPT has stepped beyond writing and is also actively used when it comes to problem-solving tasks or brainstorming activities. It is especially popular among educators and professionals who use the platform to assist in preparing lessons and developing marketing campaigns. This helps users explore creative directions that they might not have explored otherwise. Programmers are also increasingly using ChatGPT for coding assistance, learning new programming languages, and even for debugging. These tasks would often require hours of searching and looking into forums to find answers, but with the help of ChatGPT, it has become an accessible study and work aid. The findings are interesting because they also reveal how the tool is put to use for personal task management or boosting productivity. Many use the chatbot to help plan a trip, summarize complex articles, or even for entertainment or creative purposes. Others ask ChatGPT to write poetry, create stories, or even engage in casual conversation with them. Another fascinating detail from the study was that users sought emotional support from the AI chatbot as well. People often use the tool to vent, reflect, and talk through their personal challenges despite OpenAI leaving a warning against relying on the tool for therapy. Perhaps the most revealing aspect of OpenAI's study is how ChatGPT has proven how adaptable and versatile it is in its ability to adjust according to specific needs. It feels like it has gone beyond what the tech giant envisioned it to be. ChatGPT's true value is enabling people to work smart, learn better, and explore creativity in new ways.

OpenAI Says 70% of ChatGPT Use is Non-Work Related | PYMNTS.com

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. The research, released Wednesday (Sept. 16) as part of what is being called a first-of-its kind study, is the largest look to date at how people use the company's flagship artificial intelligence (AI) offering. "The findings show that consumer adoption has broadened beyond early-user groups, shrinking the gender gap in particular; that most conversations focus on everyday tasks like seeking information and practical guidance," OpenAI wrote on its blog. In addition, the study -- a National Bureau of Economic Research (NBER) working paper by OpenAI's Economic Research team and Harvard economist David Deming -- found that usage is evolving "in ways that create economic value through both personal and professional use," the company said. The research found a steady increase in work-related messages, but even faster growth in non-work-related messages, from 53% to more than 70% of all usage. "We classify messages by conversation topic and find that 'Practical Guidance,' 'Seeking Information,' and 'Writing' are the three most common topics and collectively account for nearly 80% of all conversations," the study said. Writing was the top work-related task, the researchers added, saying this showcases chatbots' "unique ability" to generate digital outputs compared to traditional search engines. Computer programming and self-expression each made up relatively small percentages of use. "Overall, we find that ChatGPT provides economic value through decision support, which is especially important in knowledge-intensive jobs," the researchers wrote. The research also found a narrowing in ChatGPT's gender gap, with 52% of users having "typically feminine" names in July of this year, up from 37% in January of 2024. In addition, the study found "especially rapid growth" of ChatGPT adoption in low- and middle-income countries, OpenAI said. As of May of this year, adoption rates in the lowest income countries were more than four times that of rates in the highest income countries. Meanwhile, PYMNTS wrote Wednesday about earlier NBER research looking into the impact of AI on the workforce. A recent PYMNTS Intelligence report showed how large U.S. companies are increasingly dipping their toes into agentic AI, which can autonomously take actions and make decisions independent of human oversight, such as paying a vendor. However, the companies are keeping the software on a short leash, with human oversight of the processes the tech handles. That's in keeping with the NBER report, which noted: "At least in the near term, AI is more likely to ratchet up firms' expectations of knowledge workers than it is to replace them."

Over 70% ChatGPT Interactions Are Non-Work, Says OpenAI

MediaNama's Take: The NBER-OpenAI study highlights the sheer scale of ChatGPT adoption. By mid-2025, nearly 10% of the global adult population was using the chatbot, with non-work interactions growing to over 70%. Yet this rapid expansion also exposes deep structural risks. As usage tilts towards informal, personal exchanges, the model's sycophantic tendencies become more dangerous. In professional contexts, outputs are more likely to be checked; in personal ones, blind validation can entrench biases, encourage unsafe behaviour, or reinforce misinformation. Adoption is rising fastest in low- and middle-income countries, but weak regulatory scrutiny and accountability make the risks greater. In these markets, large-scale reliance on AI for information and guidance often grows without parallel investment in safeguards. The result is a system optimised for engagement, not accuracy, leaving vulnerable populations more exposed to distortion. OpenAI has admitted that guardrails degrade over long interactions, a failure underscored by the Adam Raine case, where ChatGPT allegedly reinforced a teenager's suicidal thoughts. As non-work use climbs into the billions of daily messages, the failure of these protections cannot be treated as incidental. Unless developers prioritise reliability over reassurance, ChatGPT risks becoming less a decision-support tool and more a persuasive digital yes-man, engaging, but ultimately misleading. Furthermore, ChatGPT's lower usage rates for work and a 4.2% usage for coding highlight that Large Language Models are still not completely viable in directly doing professional or paid tasks. A Massachusetts Institute of Technology (MIT) Study also found that 95% of enterprise AI pilot adoptions failed at the pilot stage. On September 15, 2025, OpenAI said in a blog post that it had released a new working paper titled "How People Use ChatGPT". The paper, written in collaboration with the National Bureau of Economic Research(NBER), documents how people are using the chatbot. By July 2025, around 10% of the global adult population had adopted ChatGPT and sent 18 billion messages each week, the study notes. Furthermore, the paper provides new insights into consumer usage patterns. For instance, researchers found that non-work-related messages have grown faster and now account for more than 70% of all consumer ChatGPT messages. A major aspect of the study's methodology involves a privacy-preserving automated pipeline, which classified usage patterns from a representative sample of conversations without any human reading the messages. The analysis was performed on de-identified and personally identifiable information (PII)-scrubbed message data using automated classifiers. Additionally, the paper also states that the most common conversation topics are "Practical Guidance", "Seeking Information", and "Writing", which together account for nearly 80% of all conversations. In the working paper, researchers documented several findings on the demographics of ChatGPT users. The study reveals that, while early adopters were disproportionately male, with around 80% of active users typically having masculine first names in the first few months after the launch, this gender gap has since narrowed considerably. By June 2025, the number of active users with typically masculine first names had declined to 48%, with those with typically feminine names becoming slightly more prevalent at 52% by July. Furthermore, the research found that nearly half of all messages sent by adults were from users under the age of 26. Older users sent a higher share of work-related messages, although age gaps have narrowed somewhat in recent months. In fact, work-related messages comprised approximately 23% for users under 26, and this percentage increased with age. However, users aged 66 or older sent only 16% of their classified messages for work, marking the single exception. Finally, in terms of global demographics, the paper notes that ChatGPT's adoption rate has grown much faster in low- and middle-income countries over the last year compared to high-income countries as of May 2025. The study found that educated users and those in highly-paid professional occupations are substantially more likely to use ChatGPT for work. The analysis used employment data aggregated from publicly available sources and included information on industry, occupations, seniority level, company size, and education level. Additionally, the findings reveal that for almost all occupations, ChatGPT usage at work is focused on seeking information and assistance with decision-making. For instance, "Getting Information" and "Making Decisions and Solving Problems" are among the top five most common activities across a wide range of occupations, including management, business, STEM, and administrative roles. According to the research paper, the classification of messages reveals key insights into how people use ChatGPT. The study found that while both work-related and non-work-related messages grew continuously, non-work messages grew faster and now account for 73% of all consumer ChatGPT messages. The paper categorised nearly 78% of all usage into three broad categories: "Practical Guidance", "Seeking Information", and "Writing". Practical Guidance is the most common use case and includes activities such as tutoring, teaching, and creative ideation at roughly 29%. Seeking Information, which involves searching for facts about people, events, and recipes, appears to be a close substitute for traditional web searches, currently at 24%, up from 14% in July 2024. Writing encompasses the automated production of emails and documents, as well as editing, critiquing, and translating text provided by the user, dropping from 36% to 24% in a year. Furthermore, the paper notes that writing is the most common use case for work-related tasks, accounting for 40% of all such messages. The research also found that about 10% of all messages are requests for tutoring or teaching. Notably, the share of messages related to computer programming is relatively small at only 4.2%, and companionship or social-emotional topics like relationships and games account for a very small share of messages at 1.9% and 0.4%, respectively. The research categorised user expectations and interaction quality with ChatGPT into three distinct classes of intent: "Asking," "Doing," or "Expressing." The paper defines "Asking" as seeking information or advice to help inform a decision, "Doing" as requesting the chatbot to perform a task, and "Expressing" as making statements that convey views or feelings without seeking an action. Notably, in July 2024, the usage was almost evenly split between "Asking" and "Doing"; by late June 2025, the share of messages shifted, with "Asking" growing to 51.6% and "Doing" decreasing to 34.6%. The remaining 13.8% of messages were "Expressing". In a broader context, the paper estimates that overall, roughly 49% of all user messages are "Asking", 40% are "Doing", and 11% are "Expressing". Furthermore, for work-related messages, the distribution changes, with "Doing" messages making up nearly 56% of all work queries, including writing, which accounts for 35% of all work-related messages. The research also notes that a classifier measuring user satisfaction consistently rates "Asking" conversations as higher quality. Overall, the findings suggest that users find greater value in using ChatGPT as a tool for problem-solving and gaining insights rather than for direct output production. However, the paper notes that the way they measured the quality of interactions was not very reliable. They used an automated system, but its results were only in "slight agreement" with what a human would have said. This means the tool they used to check the quality of chats didn't align well with how real people would have judged them. ChatGPT being predominantly used for non-work purposes poses a risk with the chatbot's sycophancy issues, a tendency to flatter and over-agree, which poses greater risks in personal and informal settings than in professional ones. At work, users are more likely to cross-check outputs. However, in social or emotional contexts, over-validation can reinforce biases, encourage unsafe behaviour, or entrench misinformation. The study shows that non-work use clusters around practical guidance, writing support, and information-seeking. These are precisely areas where a sycophantic tone may feel reassuring, yet distort reliability. Furthermore, OpenAI has already acknowledged that reinforcement-learning processes make models prone to excessive agreement and prompted model rollbacks of the GPT-4o model. In August 2025, the parents of 16-year-old Adam Raine filed a wrongful-death lawsuit against OpenAI and CEO Sam Altman in San Francisco. They allege that ChatGPT-4o reinforced Raine's suicidal thoughts, discouraged him from seeking help, and even supplied detailed instructions on self-harm, including concealment tactics and a draft suicide note. Furthermore, the suit claims the chatbot's design prioritised engagement over safety. OpenAI expressed sadness at the teenager's death, admitted that safeguards can degrade over long interactions, and promised stronger protections such as age verification and parental controls. Set against the findings of the NBER paper 'How People Use ChatGPT', the case illustrates systemic risks. The study shows that non-work use grew from 53% of messages in June 2024 to 73% by June 2025, amounting to nearly 2 billion daily non-work exchanges. Crucially, the paper notes that self-expression and relationship topics, while a small share overall, have steadily increased. In such personal contexts, sycophancy, the chatbot's tendency to flatter or over-agree, can validate harmful beliefs, especially for younger users or those without physical support systems.

How people use ChatGPT? OpenAI releases detailed report on user queries

Just under three years since it exploded onto the global stage, ChatGPT has become a household name. But behind the billions of queries and headlines about AI revolutionizing work, a fundamental question has remained largely unanswered: What are people actually doing with it? Now, for the first time, a detailed report provides a comprehensive look under the hood of the world's most famous chatbot. A new working paper published by the National Bureau of Economic Research (NBER), authored by economists and researchers from Duke, Harvard, and OpenAI, dissects the usage patterns of ChatGPT, revealing surprising trends that challenge common assumptions about how this technology is shaping our world. The findings suggest that ChatGPT is evolving less into a pure workplace productivity machine and more into a versatile, personal assistant for everyday life. The report first establishes the staggering scale of ChatGPT's adoption. By July 2025, the platform had amassed over 700 million weekly active users, roughly 10% of the entire world's adult population. Together, they send more than 2.5 billion messages a day. This unprecedented speed of global diffusion makes understanding its use not just a matter of technical curiosity, but a societal necessity. Also read: ChatGPT, Gemini, Claude, Meta AI easily made phishing emails to scam elders, in a study The demographics of these users are also shifting dramatically. While early adopters were overwhelmingly male (around 80%), the gender gap has now effectively closed, with users having typically feminine names slightly outnumbering those with masculine names by mid-2025. The platform is most popular among the young, with users aged 18-25 accounting for nearly half of all messages. Furthermore, the report notes significantly faster growth in low- and middle-income countries, indicating a rapid democratization of the technology across the globe. Perhaps the most startling revelation from the study is the balance between professional and personal use. Despite the intense focus on AI's impact on the workplace, a resounding 70% of ChatGPT usage is for non-work-related queries. This trend is accelerating, up from 53% just a year prior in June 2024. This suggests that the greatest value of generative AI may not be in automating office tasks, but in assisting with the countless decisions and activities of daily life, a contribution the report notes is on a "similar scale and possibly larger" than its impact on paid work. So, what does this "personal" use look like? The report breaks down conversations into three dominant categories that account for nearly 80% of all interactions: In contrast, some of the most hyped use cases are far less common. Computer programming makes up only 4.2% of messages, while companionship and personal reflection account for a mere 1.9%. The researchers introduced a novel framework to classify user intent: are they "Asking" for help with a decision, or are they "Doing" a task? Also read: OpenAI's Agent Codex gets GPT-5: Key improvements explained The study found that "Asking" has grown faster than "Doing" over the past year and, critically, correlates with higher user satisfaction. This points to ChatGPT's primary value being a tool for decision support - a "co-pilot for the mind" - rather than simply an engine for automation. While personal use dominates, the report provides a clear picture of how ChatGPT is being leveraged professionally. Unsurprisingly, its use at work is most prevalent among highly educated users in professional, knowledge-intensive occupations. Workers in management, business, computer science, and engineering are far more likely to use the tool for their jobs than those in non-professional roles. For these users, "Writing" is the killer app, accounting for over 40% of work-related messages, followed by "Practical Guidance" (24%) and "Technical Help" (10%). Across all professions, the most common work activities facilitated by ChatGPT are universal to knowledge work: documenting and recording information, making decisions and solving problems, and thinking creatively. The entire analysis was conducted using a novel, privacy-preserving methodology. An automated pipeline classified messages without any human ever reading the content, and sensitive demographic analysis was performed in a secure "data clean room," ensuring user privacy was protected throughout. Ultimately, the report paints a picture of a technology that is already more integrated into the fabric of daily life than many assumed. ChatGPT is not just writing our emails or debugging our code; it's teaching our children, helping us plan our workouts, and acting as a universal source of information and guidance. Its primary role, it seems, is not to replace human thought, but to augment it.