Generative AI Technology

Generative AI systems are trained using a combination of supervised learning and reinforcement learning techniques. Many generative models—particularly large language models—operate by predicting the next word, symbol, or element in a sequence based on context. This prediction process is supported by training on large datasets, which helps improve accuracy, coherence, and relevance. 

During interaction, Generative AI considers the context of a user’s input and generates responses that are grammatically and semantically coherent. Reinforcement learning often incorporates human feedback to help the system better follow instructions, improve usefulness, and align outputs with expected norms. While these systems have access to a broad range of information, they can still produce factually incorrect, incomplete, or biased outputs, so it is important to verify AI-generated information before using it in academic or professional work. 

Due to their flexible design, Generative AI tools can perform tasks beyond those they were initially designed for, including: 

• Writing and revising text across genres 
• Translating content between languages 
• Generating summaries and study materials 
• Writing stories, poetry, scripts, or song lyrics 
• Creating code and assisting with debugging 
• Solving mathematical problems and explaining steps 
• Simulating conversations or role-play scenarios 
• Supporting game design or interactive simulations 

As with any technology, Generative AI is not inherently good or harmful—its impact depends on how it is used. Some concerns within the educational community focus on potential misuse, such as: 

• Generating student essays, papers, or creative work 
• Answering quiz or exam questions 
• Completing programming assignments 
• Solving math problems without student engagement 

Note: Performance varies based on the model, training data, prompt quality, and use case. All AI-generated output should be evaluated critically. 

It is also important to recognize that tools supporting these activities have existed for many years, and concerns related to academic integrity predate Generative AI. 

The following strategies can help instructors communicate expectations, design meaningful assessments, and support student learning in an AI-enabled environment. The goal is to encourage skill development, ethical use, and transparency. 

Your First Step 

A critical first step is becoming familiar with Generative AI tools, their capabilities, and their limitations. In addition to reading about them, try using these tools yourself. When experimenting, consider how your students might interact with them and how they could impact your course activities and assessments. 

Syllabus: Set Clear Policies 

Clearly outline expectations regarding the use of Generative AI tools in your course. This may include: 

• Whether AI use is permitted or restricted 
• Expectations for citation or acknowledgment 
• Consequences for inappropriate use 

Policies should align with course learning outcomes and institutional guidance. 

Discuss Expectations with Your Students 

Equally important is having open conversations with students about Generative AI. Discuss its capabilities, limitations, ethical considerations, and relevance to your discipline. Emphasize that critical thinking, evaluation of information, and digital literacy are essential skills that extend beyond the classroom and into the workforce. 

Encourage transparency by asking students to acknowledge AI use when appropriate. 

Impart a Growth Mindset 

Learning is a process that includes challenges and struggles. Help students understand that effort and persistence are part of intellectual growth. Too much reliance on AI shortcuts can limit opportunities to develop essential skills and deeper understanding. 

Assessment Strategies

Exams and Quizzes 

If exams or quizzes are used, consider avoiding high-stakes weighting. Lower-stakes assessments can reduce pressure and discourage misuse while supporting skill development. 

1. Consider Time Limits: Appropriate time limits can reduce opportunities to search for answers externally. Suggested guidelines include: 
   • 30 seconds per true/false question 
   • 60 seconds per multiple-choice question 
   • 2 minutes per short-answer question 
   • 10–15 minutes per essay question 
   • 5–10 minutes for review Adjustments may be necessary based on question complexity and accessibility needs.  
2. Online Proctoring Tools: Tools such as Lockdown browsers with monitoring features may help deter misconduct. Their effectiveness depends on consistent review of results and alignment with course goals. 

Writing Assignments 

Writing assessments may include essays, reports, reflections, or projects. 

AI detection tools may help identify content potentially generated by AI, but they are not definitive and should never be used as the sole basis for academic misconduct decisions. Human judgment and institutional policies remain essential. 

Use a plagiarism detection tool such as Turnitin to support academic integrity. It is important to review originality reports carefully, as similarity percentages may include properly cited sources and reference lists and do not automatically indicate pplagiarism. To learn more, please visit the following Turnitin support articles on How to Enable Turnitin for an Assignment. 

 Turnitin also offers an AI writing assessment to help identify text that may be AI-generated; however, these results are not always accurate and should not be used as the sole basis for academic misconduct decisions. Human judgment and institutional academic policies must guide any final determinations. To learn more, please visit the following Turnitin support articles on Using the AI Writing Detection Report and Understanding false positives within our AI writing detection capabilities. 

Using clear rubrics with well-defined criteria can help deter inappropriate AI use by clarifying expecitations and emphasizing original thinking and analysis. If you want to learn more, please visit Beyond Fairness and Consistency in Grading: The Role of Rubrics in Higher Education. 

Making Connections 

Design assignments that require students to connect course content to personal experiences, current events, or reflective analysis. These forms of learning are more difficult for AI tools to replicate meaningfully and promote deeper engagement. 

Rethinking Assessments to Include AI 

Authentic assessment design is not new, but Generative AI introduces new opportunities. Since students are likely to encounter AI tools in the workplace, instructors may consider designing activities that intentionally incorporate AI use—focusing on evaluation, reflection, critique, and ethical application rather than avoidance alone. 

Generative AI can support education in a variety of ways, including: 

• Providing formative feedback on writing 
• Serving as a supplemental FAQ or study aid 
• Supporting language learning and translation 
• Generating summaries, flashcards, or practice questions 
• Assisting with tutoring and step-by-step explanations 
• Helping create instructional materials or assessments 
• Supporting accessibility through alternative formats and summaries 

Generative AI should not replace instructors, but rather assist educators in supporting learning, providing feedback, and enhancing engagement when used thoughtfully and responsibly.