Over half of American adults have used an AI chatbot, survey finds

Artificial intelligence technology is becoming increasingly integral to everyday life, with an Elon University survey finding that 52% of U.S. adults have used AI large language models like ChatGPT, Gemini, Claude and Copilot. 

The survey, conducted in January by the Imagining the Digital Future Center at the university in North Carolina, found that 34% of its 500 respondents who had used AI said they use large language models (LLMs) at least once a day. Most popular was ChatGPT, with 72% of respondents reporting they have used it. Google’s Gemini was second, at 50%. 

It has become increasingly common for people to find themselves developing personal relationships with AI chatbots. The survey found that 38% of users said they believe LLMs will “form deep relationships with humans,” and over half reported having had spoken conversations with chatbots. Around 9% of users said the main purpose they use the models for is “social kinds of encounters like casual conversation and companionship.” The respondents found that the models can express a variety of personality traits, including confidence, curiosity and even senses of humor. 

“These findings start to establish a baseline for the way humans and AI systems will evolve together in the coming years,” Lee Rainie, director of the Imagining the Digital Future Center, told NBC News in a statement. “These tools are increasingly being integrated into daily life in sometimes quite intimate ways at the level of emotion and impact. It’s clearly shaping up as the story of another chapter in human history.”

That is consistent with an overall trend that found that 51% of respondents use LLMs for personal endeavors, rather than work-related activities.

When it comes to using the models for work purposes, respondents reported that they have used them with work-related apps such as Slack, PowerPoint and Zoom. They have also used the models to do such things as write emails, research ideas and summarize documents. Over 50% of respondents said the models have helped them improve their productivity.

Many respondents reported having anxieties about the technology. Sixty-three percent thought the models could replace a significant amount of human-to-human communication, and 59% thought they could cause a significant number of job losses.

AI technology is becoming more popular as President Donald Trump’s administration has been pushing for increased investment in AI technology. 

In January, Trump announced a joint venture among OpenAI (the company that developed ChatGPT), Oracle and SoftBank. The executives of the AI companies committed to invest $100 billion in the initiative up front and up to $500 billion over the next four years. Trump also signed an executive order about AI a few days after he entered office. 

When it comes to politics, the survey found that Republican LLM users are more likely than Democrats to use the models to search for political news and information. Thirty-four percent of respondents reported using them to get news or political information. 

Vice President JD Vance has been particularly outspoken about bias in AI, telling world leaders at an AI summit in Paris that the “Trump administration will ensure that AI systems developed in America are free from ideological bias and never restrict our citizens’ right to free speech.” The survey found, however, that 49% of users believe the creators of the LLMs should be responsible for limiting bias in LLM answers rather than government officials and lawmakers. 

As schools have also been in limbo trying to find the best way to address the use of AI technology in classrooms, the survey found that 36% of respondents use the models for schoolwork or homework. The survey also found that 77% of LLM users are ages 18 to 29. Households with children under 18 were also more likely to use LLMs, compared with those without.