ChatGPT and other artificial intelligence assistants have billions of users — but data on how people are really using the tools is sparse and fragmented.
It’s one reason why researchers launched the AI Observatory, a public resource to measure how people use AI by analyzing anonymized chats from seven different sources.
Looking at more than 23,000 conversations people had with various chatbots including OpenAI’s ChatGPT, Google’s Gemini and SpaceX’s Grok between 2023 and 2025, the researchers behind the project found that 48 percent of the chats were not about work. (The Washington Post has a content partnership with OpenAI.)
OpenAI, Google and Anthropic, maker of the Claude chatbot, all recently released reports based on user data that tried to predict how their tools may impact the future of jobs. But the observatory’s analysis showed that people are increasingly integrating AI into intimate parts of their life, like health and relationship advice, said Anka Reuel, one of the lead researchers on the project.
The findings underscore the need for independent public monitoring of how people are using AI in the real world, she said.
“I want decision-makers to look at the whole spectrum of evidence — seeing how models are actually being used across companies, across use cases, across work and non-work settings, to make informed decisions on what the policy response should be,” said Reuel, a Stanford University PhD candidate.
Catch up quickly every weekday morning with a rundown of the 7 most important and interesting stories. Sign up for The 7 newsletter.
The post New data shows people are getting personal with AI appeared first on Washington Post.
