Universities are shifting from general-purpose AI chatbots to specialized tutoring systems built directly into course materials, offering students immediate, discipline-specific help while maintaining faculty oversight of learning outcomes. South Dakota State University (SDSU) has expanded access to AI Tutor Pro across undergraduate and graduate courses, while education experts argue that the key to effective AI in classrooms lies not in replacing teachers but in creating interactive learning partnerships that guide students toward their own answers.

What Makes Course-Embedded AI Different From ChatGPT?

The distinction between a general-purpose AI chatbot and a course-embedded tutor may seem subtle, but it fundamentally changes how students learn. When a student in an SDSU history course encounters a confusing passage about the American Revolution, they no longer need to leave the course to search the internet or ask a general chatbot. Instead, they can access an AI tutor trained specifically on that course's materials.

This focused approach addresses a critical problem with broader AI tools: they can provide technically correct but contextually irrelevant answers. A course-embedded tutor, by contrast, understands the specific readings, lectures, and learning objectives students are working with. Victor Taylor, vice provost for graduate education and extended studies at SDSU, explained the practical difference: "It's not like Google where it goes to the whole internet. These are very specific AI tools that can do deep dives on information, so a student doesn't have to leave the course. They can go to the AI tutor and get specific disciplined-based responses".

"It's not like Google where it goes to the whole internet. These are very specific AI tools that can do deep dives on information, so a student doesn't have to leave the course," said Victor Taylor, vice provost for graduate education and extended studies at SDSU.

Victor Taylor, Vice Provost for Graduate Education and Extended Studies, South Dakota State University

SDSU's AI Tutor Pro has been available for more than a year and has generated positive feedback from both students and faculty across disciplines including mathematics, economics, and the humanities. The tool has been used more than 2.8 million times by students and educators in Canada and elsewhere, and SDSU offers it at no cost through a partnership.

How Should Students Actually Use AI Tutors for Better Learning?

• Ask for Guidance, Not Answers: Effective AI tutoring works when students use the tool to ask clarifying questions and request hints rather than requesting final answers. The tutor should guide reasoning by asking what step a student would try first or identifying where logic went wrong before revealing more information.
• Verify Information Against Course Materials: Students should understand that AI tutors can make mistakes and should compare important explanations with textbooks or verify them with instructors when uncertainty remains. Treating the system as an interactive study aid rather than an unquestionable source of truth produces better learning outcomes.
• Use Active Recall and Self-Explanation: Strong AI tutors encourage students to explain concepts in their own words and generate practice questions that test understanding. When a student provides an incomplete response, the tutor highlights what is missing and provides another opportunity to try.
• Engage With the Conversational Process: Rather than simply producing an answer, course-embedded tutors are designed to ask what a student already knows about a topic, assess gaps in understanding, and help guide the student toward the answer. This conversational approach matters because the goal is not for AI to complete work for students but to ensure they understand the context.

Why Universities Are Rethinking How They Assess Learning

As AI makes it easier for students to generate polished essays and solve problems in seconds, universities are reconsidering what meaningful assessment actually looks like. SDSU is piloting an emerging use of AI in teaching: oral assessment, which allows faculty to redesign assignments around oral exams rather than written work.

The tool can ask questions based on course content, and students can respond orally using a microphone or by typing responses. For a human resources course, for example, a student could play the role of an employer while the AI agent acts as a prospective employee, creating a simulated interview environment. This approach gives students a chance to explain concepts in their own words and practice spoken communication skills.

"Oral assessment is making a comeback across disciplines. AI agents can act as an 'interviewer' asking students questions about course content and engaging them in a dialogue," said Dennis Hedge, provost and vice president for academic affairs at SDSU.

Dennis Hedge, Provost and Vice President for Academic Affairs, South Dakota State University

Faculty members remain central to this process. While the AI tutor can provide feedback, grading and evaluation always remain with the professor. SDSU plans to pilot the oral assessment tool in both a STEM course and a humanities course during the school year.

What Do Experts Say About Keeping AI From Replacing Human Learning?

The broader challenge facing education is ensuring that AI enhances learning rather than shortcuts it. Adam Mangana, CEO of Optima, which operates what the company describes as the world's first virtual reality school, argues that the problem is not technology itself but passive consumption. "When kids are staring blankly at a screen, scrolling or letting it do its homework, that's when there's a problem," Mangana stated.

Adam Mangana, CEO of Optima, which operates what the company

Instead of banning AI from classrooms, Mangana suggests schools should treat it as an educational resource that acts as a personalized tutor or guide. "If we can use it as something interactive, acting as a personal tutor or a Socratic guide, it can really be a force multiplier for our students," he explained.

Mangana

"If we can use it as something interactive, acting as a personal tutor or a Socratic guide, it can really be a force multiplier for our students," said Adam Mangana, CEO of Optima.

Adam Mangana, CEO, Optima

This distinction matters because it shifts focus from what students produce to how they arrive at answers. Teachers should emphasize the process of learning rather than simply evaluating final outcomes. Mangana described AI as a potential "sparring partner" for students rather than a substitute for human thought, noting that "this tool can really maximize human potential when used properly as a sparring partner, not as something to replace humans".

The research on AI in education supports this approach. Artificial intelligence can support personalized learning paths, adaptive assessments, accessibility tools, language translation, virtual tutoring, course development, and student support. However, AI does not automatically improve learning simply because it makes tasks faster. A student who asks AI to complete every assignment may produce more work without developing deeper understanding.

How Can Schools Build Trust With Parents About AI in Classrooms?

Parents understandably worry about introducing more technology into children's lives, especially when screen time is already a concern. Mangana acknowledged this anxiety as valid but argued that schools can address it through transparency and intentionality. "They should demand transparency and intentionality," he said, emphasizing that schools need to clearly explain why they are introducing a particular technology and what educational purpose it serves.

Mangana

The goal should not be to increase screen time for its own sake. "Education shouldn't be about more screen time just for the sake of screen time. It should be about better outcomes," Mangana noted. When schools are transparent and purposeful with these tools, they can become powerful learning multipliers that provide individualized education to more students regardless of geographic location or access to traditional tutoring services.

SDSU's broader strategy reflects this philosophy. The university sees AI as a technology that serves everyone regardless of major and is integrating AI competencies throughout all academic program curricula. Dennis Hedge, the provost, emphasized that the university's approach is rooted in awareness, ethical use, and adaptability: "Instilling a sense of lifelong learning related to AI and adjusting as technology advances is key to ensuring a strong and nimble workforce for the state".

As AI becomes increasingly accessible to students both inside and outside school, the challenge for educators is not to keep the technology out of classrooms but to teach students how to use it effectively, ethically, and in ways that enhance rather than replace genuine learning.