More than a year after the California State University system invested millions in a plan to grow the state’s artificial intelligence–ready workforce, new data shows that the majority of students, faculty and staff at the nation’s largest university system are using AI-powered tools—despite some reluctance and skepticism.
ChatGPT, which the entire CSU system has access to through a partnership with OpenAI, is by far the most popular tool.
On Wednesday, the CSU system—which has 460,000 students and 63,000 faculty and staff across 23 campuses—published a report detailing those and other findings from a fall 2025 survey of 94,060 respondents—including 80,626 students, 6,085 faculty and 7,349 staff—about how they’re using AI.
According to the survey, the vast majority of respondents (95 percent) have used at least one of the 21 AI tools listed in the survey. Additionally, 64 percent of staff, 59 percent of faculty and 53 percent of students report using it on a consistent, ongoing basis—including many who also report using AI tools outside of work.
“This survey captures a moment of transition in higher education, where both students and faculty are actively assessing how AI fits into teaching and learning,” David Goldberg, an AI faculty fellow at San Diego State University and one of the lead researchers on the survey, said in a news release. “The data gives us a powerful foundation to better support faculty by tailoring training to real needs, bringing more consistency to AI use in the classroom, and ensuring that its use strengthens learning outcomes. It also offers a roadmap for institutions nationwide to better understand AI’s role and to implement it thoughtfully, consistently, and responsibly.”
Despite the results showing the CSU system’s embrace of AI, the data also suggests that faculty haven’t come to a consensus about its place in the classroom.
Just over a quarter of faculty respondents said they remain neutral on student use of AI; 22 percent discourage it while 19 percent encourage it. Just under one in five (18 percent) forbid it entirely, but 6 percent require it and 7 percent don’t address it with their students at all. At the same time, over half (55 percent) of faculty are using AI to develop course materials, while 69 percent provide students with guidance for using it effectively and responsibly.
Cal State Is ChatGPT Country
While there are numerous AI-powered tools on the market, 84 percent of students, 87 percent of faculty and 89 percent of staff ranked ChatGPT as their most used tool. Other tools—including Grammarly, Canva, Microsoft Copilot and Google Gemini—varied in popularity and use.
Although ChatGPT became the first mainstream large language model when OpenAI debuted it in November 2022, evidence of its widespread use at the CSU system comes after the institution signed a $17 million contract with OpenAI last year to provide everyone on campus access to ChatGPT Edu. The advantage of paying for the education-specific version of an LLM, according to OpenAI, is that it can be used within a closed, secure system. While other tech companies have unrolled similar products, OpenAI is so far dominating the higher education market. As of December, it had sold more than 700,000 ChatGPT licenses to at least 35 public universities.
However, some faculty at the CSU system have pushed back against the OpenAI contract, asserting that there’s not much evidence that ChatGPT Edu benefits teaching and learning. “It is a general-purpose chatbot that is not designed, trained, or optimized for education,” reads a faculty-led petition calling on Chancellor Mildred García not to renew CSU’s contract with OpenAI, which expires June 30, and instead reinvest the money in human labor. “Beyond its privacy and security features, ChatGPT Edu is identical to the free online version of ChatGPT,” it said.
Although the survey doesn’t differentiate between usage of ChatGPT and ChatGPT Edu, it found that more than half of faculty, one-third of staff and one-fourth of students access AI tools through their institutions.
In the survey, faculty also emerged as the leading critics of AI.
“I worry we’re training students to be dependent on AI rather than developing their own critical thinking skills,” one faculty respondent said. “I grapple with preparing students for this inevitable technology without endorsing the makers of the tools who are advancing carelessly and without regard for societal harms,” said another.
Still, faculty were split about the effect AI has had on their teaching—56 percent reported a positive effect, while 52 percent said the impact has been negative. By comparison, 72 percent of staff and 64 percent of students reported a positive effect, but about 80 percent of students said they felt uncomfortable submitting AI-generated work as their own. The majority of all respondents said verifying the accuracy of AI-generated content is necessary.
Separate survey data from Educause shows that faculty and staff are using AI for tasks such as brainstorming, drafting emails, summarizing long documents or meetings, proofreading, and creating presentations. Students, on the other hand, use the technology for getting answers to problems, proofreading or editing their work, and summarizing lecture notes and articles.
Skepticism Persists
Despite increased adoption, 65 percent of students and 59 percent of faculty who responded to the CSU survey expressed skepticism about whether AI is benefiting education over all.
Britt Paris, chair of the ad hoc AI and academic professions committee for the American Association of University Professors, told Inside Higher Ed that the CSU system’s survey results show that “people in higher ed are becoming more critical of AI as it relates to the work of teaching and learning,” and that “corporate AI’s intervention into education has given instructors and learners pause.”
Mixed feelings aside, the survey showed that most people across the CSU system are feeling the pressure to jump on the AI bandwagon amid the tech sector’s ongoing predictions that it will reshape—and shrink—the entry-level job market in the years to come.
About 82 percent of staff, 78 percent of faculty and 69 percent of students said they believe AI will become an essential part of most professions. Meanwhile, a similar share expressed concern about job security. The majority of staff (82 percent) and faculty (72 percent) also want formal AI training, whereas nearly half of students (49 percent) said the same.
“Even though I don’t want to use it, I HAVE TO!” one student respondent said. “Because if I don’t, then I’ll be left behind and that is the last thing someone would want in this stupid job market.”
