Amid Google’s search supremacy, OpenAI has said that a growing number of users are turning to ChatGPT to look up information online.
The Microsoft-backed Artificial Intelligence (AI) startup reported that as of July 2025, 24 per cent of user conversations with ChatGPT were focused on seeking information – up by 10 per cent from a year ago. These figures were released as part of a broader, 64-page National Bureau of Economic Research (NBER) working paper authored by OpenAI’s Economic Research team and Harvard economist David Deming.
The study is aimed at tracking how consumer usage of ChatGPT has evolved since its global debut roughly three years ago. Asking ChatGPT for practical guidance, information-seeking queries, and writing-related tasks account for nearly 77 per cent of all user conversations, as per the paper.
Since its launch in 2022, ChatGPT has seen explosive growth with the consumer AI chatbot app surpassing 700 million weekly users or 10 per cent of the global population. The first-of-its-kind working paper by OpenAI not only offers a window into how consumers are using the world’s most popular AI chatbot but also provides insights into daily user interactions with generative AI. It also carries potential implications for various industries and sectors such as software development, e-commerce, healthcare, etc.
“Together, these findings from the largest study of ChatGPT consumer usage to date show not only who is using AI and what they’re using it for, but also how it is creating real economic value that is increasingly central to people’s work and everyday lives,” OpenAI said in a blog post on Monday, September 15.
As part of the study, researchers analysed over 1.5 million user conversations from May 2024 to July 2025 in a privacy-preserving manner. OpenAI has emphasised that the researchers were not privy to messages sent by users as they relied on automated tools to detect usage patterns.
The sampling dataset of the study excluded conversations from under-18 users, deactivated or banned users, logged-out users, users who have deleted their ChatGPT accounts, and users who had opted out of sharing their messages for AI model training.
ChatGPT Enterprise users’ conversations were also excluded as the study only focused on consumer plans like ChatGPT Pro and ChatGPT Plus as well as the free tier.
The paper noted that only 30 per cent of consumer usage of ChatGPT is work-related, highlighting how enterprise adoption of AI is still in the nascent stage globally. A recent study conducted by Massachusetts Institute of Technology (MIT) found that 95 per cent of US-based firms that had invested somewhere between $35 billion to $40 billion in generative AI, saw little to no returns largely due to flawed enterprise integration of AI.
OpenAI acknowledged that non-work-related messages are growing faster, rising from 53 per cent to 70 per cent of all ChatGPT usage. It also pointed out that use of ChatGPT for work-related purposes is more common among educated users in highly paid professional occupations. The share of work-related ChatGPT messages was also higher among younger users.
“Work-related messages comprised approximately 23 per cent of messages for users under age 26, with this share increasing with age. The one exception is users who self-attest to being 66 years-old or older, with only 16 per cent of their classified messages being work-related,” the study said.
Since ChatGPT’s debut, a key question posed by enterprise leaders has been: how much economic value can be unlocked using AI and whether it can boost productivity by measurable amounts within an organisation.
Dismissing GDP contribution as an effective way to measure the impact of AI, the study noted that ChatGPT is creating value by providing support for decisions taken by users. “ChatGPT helps improve judgment and productivity, especially in knowledge-intensive jobs. And as people discover these and other benefits, usage deepens—with user cohorts increasing their activity over time through improved models and new use-case discovery,” OpenAI said.
OpenAI’s analyses were based on conversations sampled from a subset of 1,30,000 ChatGPT Free, Plus, and Pro users that were aggregated by a vendor working through a secure Data Clean Room (DCR).
– Rise of AI coding tools: The share of users turning to ChatGPT for technical help, such as assistance in coding, declined from 12 per cent to 5 per cent between July 2024 to July 2025. OpenAI attributed the drop to the growing use of its Application Programming Interface or API (outside of ChatGPT) for programming, as well as the emergence of autonomous programming agents like Codex.
– Impact of viral trends: The use of ChatGPT for generating images and other multimedia only grew by 5 per cent with a large spike recorded in April this year, owing to the Ghibli-style AI images trend that went viral on social media.
– Use of chatbots for therapy: About 49 per cent of messages to ChatGPT involved ‘Asking’ prompts, indicating that the chatbot is valued more as an advisor rather than for completing tasks such as drafting emails. However, only 11 per cent of messages involved personal reflection, exploration, and play, as per the study.
– Countries leading AI adoption: As of May 2025, ChatGPT adoption growth rates in the lowest-income countries were four times more than those in the highest-income countries.
– Shrinking gender gap: The study found that ChatGPT’s gender gap had been narrowed dramatically, with 52 per cent of users having typically feminine names, up from 37 per cent in January 2024.
Around 46 per cent of messages were sent to ChatGPT by users between 18 to 25 years of age, as per the study. However, they only include individuals who self-reported their age while signing up to use the chatbot.