The University of Missouri is testing Show-Me AI and other artificial intelligence initiatives
By the 2026 academic year, the University of Missouri AI Standing Committee hopes to introduce Show-Me AI, a software that will provide students and instructors access to several generative AI tools, including ChatGPT. Show-Me AI is a walled garden tool, storing and encrypting material input by users in private servers, which allows faculty to upload students’ work for grading or AI tutor development without violating student intellectual property law.
The Family Educational Rights and Privacy Act is a federal law protecting students’ rights to control the release of their educational records, which includes files or intellectual property created by or pertaining to the student in any way. Kevin Brown, chair of the AI standing committee, explained that Show-Me AI will also allow students to input their ideas and work into Show-Me AI without feeding into a generative AI model.
“When people use free tools on the internet, and they put in their own work, they are essentially handing over that work to help train the model and to be part of the property of that company,” Brown said.
Justin Palozola, a senior technology resource manager and a leader on the Show-Me AI team, explained that when users input a prompt into a generative AI platform such as ChatGPT, that data is then sent to ChatGPT’s servers and used to train the AI model. According to Palozola, data input through Show-Me AI’s software will instead be protected through contracts with Amazon Web Services.
“The only people that have access — and it wouldn’t even be us on the admin team or anything — it would simply be whoever created that assistant and then whoever they shared that with,” Palozola said.
The Show-Me AI team hopes instructors will eventually be able to use the software to develop custom teaching assistants or tutors from the material of individual classes. The software will be introduced in waves to about 1,000 registered pilot users.
The code used to develop the software is open-sourced and was originally developed by Vanderbilt University. Vanderbilt’s name for the software is Amplify, which Mizzou renamed to Show-Me AI.
Thus, as Vanderbilt releases updates to Amplify, Mizzou is also able to apply those updates to Show-Me AI.
The first wave of approximately 55 pilot users, which consisted mainly of faculty with AI experience, gained access to the software in September. The second wave, which consisted of faculty members planning to use Show-Me AI in their spring semester classes, was set to gain access once version 8 of Amplify was released.
The Show-Me AI team expected the upgrade by the end of October or early November, but it did not occur until the week before Thanksgiving. The existing users gained access on Nov. 20, and the orientation for the new wave of users took place on Nov. 25.
Student pilot users will gain access in January. Pilot users will be able to share their feedback through discussion groups, surveys and monthly virtual meetings.
As Show-Me AI makes an appearance in classes, Mizzou’s administration also intends to introduce an AI certificate program in the spring 2026 semester.
There will be four required classes to earn the certificate: AI Essentials, Writing with AI, AI Design Studio and an elective AI course. These first two courses are currently being offered as topic classes and will be offered simultaneously during the spring semester. However, Brown said the other two courses will only be offered if enough students enroll in the first two classes.
The College of Engineering and College of Arts and Science will facilitate this program, but students of any major can get the certificate. According to Brown, there will be creative opportunities for students completing the certificate who are interested in the arts.
Brown said that students will have a special opportunity to work in a collaborative environment, gaining both engineering and humanistic perspectives through the AI program.
“We’re realizing that AI works best when it’s collaborative and not just doing things for you,” Brown said.
Palozola believes Show-Me AI will be available to all faculty and staff by the 2026-2027 academic year. However, this largely depends on the budget decisions the University of Missouri makes. The current pace of pilot testing makes the viability of this timeline unclear.
“Currently it’s a proof of concept,” said Hayden Hawkins, an undergraduate Missouri Students Association representative on the AI Standing Committee. “So, depending on how it goes this semester, (the university) could cancel it.”
According to Mizzou Deputy Spokesperson Travis Zimpfer, the current budget for Show-Me AI is $100,000. Palozola said the Show-Me AI team plans to produce a report in February or March on the initiative’s success to request more funding from the university.
Costs for the software are based on the token usage data Vanderbilt gathered from its pilot class of 1,300 users in April of 2025. Vanderbilt found that each user spent an average of $3 in token usage per month.
Tokens are units of text that are input into a large language model. Each unit can be made up of chunks of words, parts of words or characters. Simpler words may be worth one token or a portion of a token, while more complicated words may equate to multiple tokens.
“So (the model) takes your prompt, it takes however many tokens you’re using, it then multiplies it by however much that company charges,” Palozola said. “And then that’s what gets charged to us.”
Show-Me AI pilot users have a budget of $5 worth of tokens each month that they can keep track of.
Hawkins stated that he did not expect to see campus-wide access by the 2026-27 academic year, but instead more limited access within classrooms.
“I don’t necessarily see, like, everyone having full open-ended access, especially because we haven’t even figured out cost associated with it,” Hawkins said.
Artificial intelligence has been shown to have a significant impact on youth mental health.
According to the American Psychological Association, adolescents “may struggle to distinguish between the simulated empathy of an AI chatbot or companion and genuine human understanding” and thus may form unhealthy dependencies on artificial interaction. The APA recommends that AI developers “prioritize features that prevent exploitation, manipulation, and the erosion of real-world relationships.”
Currently, building specialized safeguards for Show-Me AI may be beyond the budget and resources allocated to the project by the University of Missouri.
Brown explained the Show-Me AI team would not build these safeguards themselves and will instead rely on safeguards designed by AI companies whose platforms are included in the software.
“We’ve actually talked to the Counseling Center and worked to do what we can to make sure that, you know, we can put whatever safeguards in place we can,” Brown said.
Before accessing Show-Me AI, students will also complete an orientation that’ll educate them on the proper uses of the walled garden. According to Brown, Mizzou Mental Health Services is advising the Show-Me AI technical team to ensure they are knowledgeable about student well-being.
Palozola explained that Show-Me AI does not have a system to flag entries related to mental health issues or inappropriate content. Palozola clarified that the university itself will not have access to or be able to monitor data input into the software by its users, but the users themselves will have access.
“If you ask it how to build a bomb, it’s not going to give you that information,” Brown said. “(AI companies) made similar sort of safeguards and guardrails where if, you know, someone repeatedly asks about health issues, it will refer them to mental health experts, and it will, you know, not continue to give them advice.”
Elizabeth Lucas, Mizzou Journalism School AI Fellow and professor, noted that if Show-Me AI is opened on a broad scale, then some safeguards should be put in place to flag potentially problematic conversations. Lucas was a member of the first wave of Show-Me AI pilot users.
“The fear I have around addiction with AI is the addiction to having a conversation with something that just wants to give you an answer, or please you,” Lucas said. “It is designed to respond in an appealing way. Which makes sense, these are built by people who are building a product that is only gonna make them money if people use it, and the natural logic behind that is that you want people to feel good when they use it.”
Lucas emphasized the importance of educating college students about responsible AI use instead of relying on guardrails to prevent negative impacts on their mental health.
“What kind of training are we giving students to understand AI bias?” Lucas said. “What kind of training are we giving students to understand the ethical or the environmental consequences of the use of AI? Like, I’m not saying don’t use it, but we have to talk about it holistically.”
Before the university began developing its AI initiatives, the AI Standing Committee outlined a goal to create “clear, transparent expectations around AI use in learning environments to support academic integrity and reduce confusion among students.” Committee members are revising these policies as they review students’ usage and opinions regarding AI.
While the committee established a campus-wide policy requiring instructors to include an AI policy statement in their syllabus, some confusion and anxiety remain among students on how they are allowed to use AI on different kinds of assignments.
“A lot of students are confused about how they can or can’t use AI, and there’s no great resource for, like, tracking AI usage in a class,” Hawkins said.
Brown said the committee is considering a per-assignment requirement for instructions on how a student can use AI.
Hawkins also highlighted the lack of enforceable measures the university can take against improper AI usage. The unauthorized use of artificially generated content is considered academic dishonesty and prohibited under the University of Missouri’s Standard of Conduct. The language doesn’t make it fully clear to students what unauthorized use means in practice.
Discussions in AI Standing Committee meetings also mentioned allowing students to opt out of using AI for assignments that may require AI or be AI-driven.
According to Hawkins, there is a split among student perspectives regarding AI integration in education. Some are more willing to embrace AI for classroom operations, homework and studying, while others would prefer not to use it at all.
Another reason students may avoid using AI is its environmental toll, as generative AI platforms are known for having an immense energy tax. According to MIT News, powering these platforms can cause pollution and other harmful consequences.
Lucas noted that a good amount of uncertainty still surrounds AI usage, and it is normal for people to feel overwhelmed or have questions.
“Nobody has the answers,” Lucas said. “And so everybody is just in this state of, like, trying to catch up, and, like, trying to figure out how it’s useful, and it’s not as useful as some of the hype would lead us to believe. It’s just not.”
The following editors contributed to the editing process of this investigation:
Nico Berlin | [email protected]
Avery Copeland, Emma Harper and Ava Mohror | [email protected], [email protected], [email protected]
Alex Gribb | [email protected]
Chase Pray | [email protected]