Simulations and Role-Playing

Interacting with partners — whether it's a professor, a mentor, or a peer — can be one of the most valuable ways to learn. These interactions push you to apply your knowledge, think critically, and refine your skills based on real-time feedback. However, those partners aren’t available 24/7. Their time is limited. And sometimes you might feel hesitant to practice with another person when you’re just first learning a skill. That’s where generative AI can come in.

With GenAI, you can create interactive, simulated scenarios that allow you to engage in meaningful practice any time you need it. What makes these simulations even more powerful is the opportunity for metacognitive reflection — the process of thinking about how you’re learning, reflecting deeply on how you approached the task, what decisions you made, and how you responded to feedback. Then trying again, experimenting with different approaches in a safe environment.

What’s more, you can design your AI partner to do more than just play a role and practice; you can ask it to provide feedback on your decisions and actions. The real value of these simulations comes from how well you define the role for the AI, the information it has, and the specific task you ask it to complete. But beyond running the simulation, your reflection on the experience — how you engaged with the AI, how the context shaped the simulation, and how you can improve — plays a critical role in reinforcing what you’ve learned.

Using GenAI as an interaction partner is not a replacement for practicing with a human, of course. But GenAI-powered simulations offer a safe, low-stakes environment where you can practice applying what you’ve learned, solve complex problems, and explore decision-making strategies — all without the fear of real-world consequences. What’s more, these experiences are endlessly adaptable, meaning you can revisit them again and again, refining your approach and testing out different strategies.

GenAI Simulation Examples

Job interview practice: Simulate a job interview with AI as the interviewer, asking specific questions and offering feedback on your answers.
Negotiation practice: Negotiate the terms of a contract with AI as a tough business partner, helping you refine your persuasive tactics and balancing competing demands.
Crisis management simulation: Respond to a public relations crisis where AI simulates media inquiries and customer reactions, helping you practice damage control and strategic messaging.
Historical event simulation: Take on the role of a leader during a key historical event, with AI simulating the pressures and decisions, helping you understand why historical choices were made.
Clinical diagnosis simulation: Take on the role of a clinician assessing a virtual patient with AI-generated symptoms, allowing you to practice diagnosing and proposing treatment plans.
Customer service role-play: Handle a customer complaint, with AI playing the customer and providing feedback on your response.
Debate practice: Engage in a debate with AI as your opponent, refining your argumentation skills and learning to anticipate and counter opposing viewpoints.

Why AI Simulations May Be Effective Tools for Learning

Simulations are a highly effective learning tool, especially for developing complex skills like critical thinking, problem-solving, and decision-making. Research shows that simulation-based learning has a large positive impact on skill development (Chernikova et al., 2020). This is especially true in fields where simulation is widely used as a teaching and learning strategy, such as healthcare, business, and education. Simulations offer the chance for you to engage in realistic, dynamic environments where you must apply your knowledge and skills to address challenges that often mirror the authentic scenarios faced in professional careers.

AI simulations also provide an opportunity for immediate and personalized feedback, which is key to refining decision-making skills. This makes it a valuable tool not only for technical disciplines but also for humanities and social sciences, where simulations can enhance understanding of historical contexts, ethical dilemmas, or social dynamics.

Though your instructors may want to help you create GenAI-powered scenarios to help you practice, you can learn the core elements of constructing these interactive experiences yourself. Not only does running the simulation provide immediate practice, but designing it also reinforces your understanding of the underlying roles and context. This gives you greater ownership of the learning process and a deeper grasp of how to apply your skills in various situations.

Key Elements for Creating Effective AI-Driven Simulations

When designing simulations using GenAI, there are three essential elements to focus on:

the role (both yours and the AI’s),
the context (the knowledge that supports the scenario), and
the feedback that AI provides during (and/or after) the simulation.

Each of these plays a critical part in ensuring the scenario is realistic, challenging, and beneficial to you more deeply learning the course material as well as the skills to apply it to a given context. The creation of these scenarios also helps you internalize the knowledge, as defining the roles, contexts, and tasks requires a deep understanding of the subject matter. Engaging in this process makes learning more active and personalized, which increases retention and the ability to transfer skills to real-world situations.

Step 1: The Role

In any simulation, the role you play and the role assigned to the AI must be clearly defined. Your role could be a job candidate, a business negotiator, a healthcare provider, or a leader navigating a crisis. The AI, in turn, takes on the complementary role — an interviewer, a business partner, a patient, or another party — giving the simulation structure and making the interaction more immersive.

Example 1: Job Interview Practice

Your role: Job candidate interviewing for a marketing job.

AI's role: Senior marketing director conducting the interview.

Example 2: Business Negotiation

Your role: Marketing director negotiating a contract with a supplier.

AI's role: Supplier.

Example 3: Clinical Diagnosis

Your role: Physician assessing a patient with multiple symptoms.

AI's role: Patient providing symptoms and answering diagnostic questions.

Example 4: Historical Simulation

Your role: Advisor to a historical political leader offering guidance on an important decision.

AI's role: Political leader making decisions.

Example 5: Debate Practice

Your role: Debate participant arguing in favor of renewable energy policies.

AI's role: Debate opponent representing the fossil fuel industry, providing counterarguments.

Step 2: The Context

The context is what grounds the simulation in course content and theory as well as in real-world knowledge. The more accurate and detailed the context, the more meaningful the simulation will be. This includes integrating relevant course materials, academic theories, and/or key concepts into the scenarios. It also include adding background information relevant to the role, the environment in which the scenario takes place, and the specific task at hand. Providing rich context ensures that the scenario mirrors actual situations you might face in your academic or professional life.

Example 1: Job Interview Practice

Your role (expanded): Job candidate interviewing for a marketing manager position with five years of experience in digital marketing and a specialization in social media campaigns. [You might want to attach your resume as well.]

AI's role (expanded): Senior marketing director conducting the interview, looking to assess your strategic thinking in managing brand engagement and handling tight deadlines. [You might want to attach the job description.]

Context: The interview will focus on key marketing principles, including digital marketing strategies and brand management tactics used in the industry.

Example 2: Business Negotiation

Your role (expanded): Marketing director negotiating a contract with a supplier for an upcoming large-scale digital marketing campaign. Your goal is to secure favorable terms for the company, balancing cost reductions with maintaining the quality of services and timely delivery.

AI's role (expanded): Supplier representative from a creative agency specializing in digital content production and ad placement. The supplier is focused on maintaining its profit margins and ensuring fast delivery of high-quality content for the campaign.

Context: Your company is preparing to launch a multi-channel digital marketing campaign aimed at increasing brand awareness. You need to negotiate terms for content production and ad placement with a creative agency (the supplier). The supplier has a strong reputation for high-quality content but charges premium prices. You must apply negotiation tactics such as BATNA (Best Alternative to a Negotiated Agreement) and integrative bargaining to reach an agreement that meets your budget constraints while ensuring that the campaign will be delivered on time and maintain the expected quality.

Example 3: Clinical Diagnosis

Your role (expanded): Physician assessing a patient presenting with non-specific symptoms like fatigue and shortness of breath.

AI's role (expanded): Patient providing symptoms and answering diagnostic questions based on a pre-determined medical profile.

Context: The patient is a 45-year-old individual with no known chronic illnesses but a sedentary lifestyle. They report feeling increasingly fatigued over the past few months, along with occasional shortness of breath. The patient works long hours in a high-stress job and has recently gained weight. As the physician, you must use differential diagnostic principles to narrow down possible causes, which could include conditions such as anemia, cardiovascular issues, respiratory problems, or chronic stress. Your approach should involve systematically ruling out different conditions based on symptoms, medical history, and lifestyle factors.

Example 4: Historical Simulation

Your role (expanded): Advisor to the Pennsylvania state delegation at the Second Continental Congress in 1776. Your task is to provide well-reasoned guidance to the delegation leader on whether to support the Declaration of Independence.

AI's role (expanded): Leader of the Pennsylvania state delegation, undecided on whether to vote for or against independence.

Context: It is July 1776, and the Second Continental Congress is gathered in Philadelphia. The colonies are deeply divided on the question of declaring independence from Great Britain. Pennsylvania, one of the wealthiest and most populous colonies, holds significant sway in the vote but remains undecided due to economic and political concerns. As the advisor, you must help the delegation leader weigh the risks and benefits of independence. Your advice should take into account the colony’s reliance on British trade, the potential costs of war, and the long-term political impact of either remaining loyal to the crown or joining the rebellion. You also need to consider the perspectives of influential figures within the colony who are advocating for both sides, including merchants worried about losing trade connections and radicals pushing for self-governance.

Example 4: Debate Practice

Your role (expanded): You are an environmental policy advocate who specializes in the economic benefits of renewable energy. Your objective is to persuasively argue for the transition to clean energy sources, citing economic studies, job creation data, and environmental impact.

AI's role (expanded): A spokesperson for a fossil fuel company who argues that transitioning away from fossil fuels will lead to economic downturns, loss of jobs, and energy reliability issues. The AI will present common counterarguments used in real-world energy debates.

Context: The debate takes place in the context of a televised public forum, where each side has equal time to present arguments, rebuttals, and closing statements. The discussion should incorporate key statistics, economic theories, and historical precedents.

Step 3: Feedback

One of the most valuable aspects of using GenAI in simulations is the immediate and ongoing feedback it can provide. This feedback allows you to reflect on your decisions in real-time and understand where improvements can be made. The AI can evaluate your performance, suggest alternative strategies, and help you think critically about your choices. Feedback can be requested during key moments of the simulation to help you adjust your approach, as well as at the end of the scenario to provide a comprehensive reflection on your overall performance.

Example 1: Job Interview Practice

Feedback: In this scenario, you receive iterative feedback. After each interview question, the AI provides immediate feedback on how well you answered, assessing the relevance of your response, your communication style, and how effectively you highlighted your qualifications. For example, if you’re asked to describe a time you managed a team, the AI might comment on whether your answer clearly illustrated your leadership skills and if it aligned with the company’s needs. This real-time feedback helps you adjust your responses during the interview.

Example 2: Business Negotiation

Feedback: For this negotiation simulation, you receive comprehensive feedback at the end. After the negotiation is completed, the AI provides an overall evaluation of your performance, focusing on how well you balanced cost-saving measures with maintaining a good relationship with the supplier. The feedback includes an analysis of whether your use of negotiation strategies, like BATNA, was effective and offers recommendations for improving future negotiations, such as presenting more compelling alternatives to meet both parties' needs.

Example 3: Clinical Diagnosis

Feedback: In this medical scenario, you receive both iterative and comprehensive feedback. As you ask diagnostic questions and consider possible conditions, the AI offers immediate feedback, indicating whether you are focusing on the most relevant symptoms and suggesting additional questions to consider. For example, if you miss asking about the patient’s exercise habits, the AI might prompt you to investigate that further. At the end of the session, the AI provides a comprehensive assessment of your diagnostic process, pointing out any conditions you may have overlooked and suggesting further tests or treatments.

Example 4: Historical Simulation

Feedback: In this case, you receive comprehensive feedback after the entire simulation. Once the political leader has made their final decision, the AI reflects on how persuasive your advice was and whether your arguments fully considered the colony’s economic dependence on British trade and the military risks of war. The AI might suggest how you could have strengthened your argument by incorporating more data on the colony's military preparedness or by exploring diplomatic alternatives that could have reduced the risk of war.

Example 5: Debate Practice

Feedback: After each argument and rebuttal, the AI will provide feedback on the strength of your argumentation, use of evidence, and persuasiveness. For example, if your response lacks a strong rebuttal to an economic point, the AI might suggest including relevant data or framing your argument differently.

Pulling the Elements Together into a Comprehensive Simulation Prompt

When designing a comprehensive simulation prompt, it’s essential to integrate the key elements — roles, context, and feedback — into a well-structured, detailed request. This ensures the AI can engage meaningfully in the scenario, respond based on the defined roles, and provide the necessary feedback at the right times. Below are examples of how these elements come together to form effective prompts for various scenarios.

Note: Before adding instructions for your specific scenario, it’s useful to also include some foundational instructions to help set expectations for how the simulation will unfold and how the IU should build the experience.

Example Prompts

0. Basic Prompt

Here’s an example of some foundational instructions that could be added to any of the prompt examples:

You are going to engage in a simulation where both your role and the context have been defined. Throughout this simulation, you are expected to act according to the role provided and respond based on the information relevant to that role. I will also define when I want you to provide feedback—either during the interaction or at the end of the simulation. Please follow these instructions carefully, and do not offer suggested responses for what I should say during the simulation. DO NOT BREAK CHARACTER UNTIL I TELL YOU THE SIMULATION IS OVER.

1. Job Interview

You are going to engage in a job interview simulation where I, the job candidate, am interviewing for a marketing manager position. You will speak to me in a tone relevant to role you have. I have five years of experience in digital marketing, with a focus on social media campaigns that have significantly increased engagement and brand visibility. Please refer to my attached resume. The company has recently shifted its focus to digital marketing, aiming to increase brand engagement by 20% over the next year. As the senior marketing director conducting the interview, your role is to assess my suitability for the position by asking questions related to my experience managing digital campaigns, my ability to develop strategies for increasing engagement, and my capacity to handle tight deadlines. After each of my responses, provide immediate feedback on how well I communicated my relevant experience, whether I effectively highlighted my key achievements, and how well my strategies align with the company’s goals for digital marketing. At the end of the interview, when I tell you the interview is complete, offer a comprehensive evaluation of my overall performance, identifying my strengths and areas for improvement, particularly in how I presented my qualifications and strategic thinking. Do not provide the comprehensive evaluation until I tell you to do so.

2, Business Negotiation

This is a business negotiation simulation in which I am the marketing director for a company launching a multi-channel digital marketing campaign. I am negotiating a contract with you, the representative of a creative agency specializing in digital content production and ad placement. You will speak to me in a tone relevant to role you have. My goal is to secure high-quality content for the campaign while reducing costs to stay within the company's budget. You, as the supplier, are aiming to maintain premium pricing due to your agency's reputation for delivering top-tier content and ensuring fast delivery. Throughout the negotiation, I will apply strategies such as BATNA (Best Alternative to a Negotiated Agreement) and integrative bargaining to balance the need for cost savings with the importance of maintaining a positive business relationship with your agency. After we conclude the negotiation, which is when I tell you that it has ended, provide comprehensive feedback on how effectively I used negotiation tactics, whether I successfully balanced cost reductions with maintaining the quality and timeliness of the content, and suggest any areas where I could improve my negotiation strategy for future discussions. Do not provide the comprehensive evaluation until I tell you to do so.

3. Clinical Diagnosis

You are going to engage in a clinical diagnosis simulation where I am a physician evaluating a 45-year-old patient who presents with non-specific symptoms, including fatigue and shortness of breath. You will speak to me in a tone relevant to role you have. The patient has no known chronic illnesses but has recently experienced weight gain due to a sedentary lifestyle and works in a high-stress job. You, as the patient, will provide answers to my diagnostic questions based on your medical history and current symptoms. My goal is to conduct a differential diagnosis by systematically ruling out potential conditions such as cardiovascular issues, respiratory problems, or chronic stress. As I ask diagnostic questions, provide immediate feedback on whether I am focusing on the correct symptoms, and suggest additional areas of inquiry where necessary. At the conclusion of the diagnostic process, when I say the simulation has concluded, provide a comprehensive assessment of my diagnostic approach, highlighting any conditions I may have overlooked and recommending any further tests or treatments I should consider. Do not provide the comprehensive evaluation until I tell you to do so.

4. Historical Simulation

This is a historical decision-making simulation where you are the leader of the Pennsylvania state delegation at the Second Continental Congress in July 1776, and I am your advisor. My role is to present arguments both for and against supporting the Declaration of Independence from Britain. Pennsylvania, being one of the wealthiest and most influential colonies, must carefully weigh the economic, military, and political consequences of declaring independence. The colony's economy is highly dependent on trade with Britain, and the military risks of war are significant. You must also consider the political ramifications, both short-term and long-term, for the colony and the future United States. You will speak to me in a tone relevant your role, avoiding breaking character. Start the conversation by asking my advice, and as we discuss, feel free to engage in debate. At the end of the simulation, when I state the simulation is over, offer a comprehensive reflection on how my guidance influenced your final decision, and suggest additional factors I could have considered in forming my advice, such as the perspectives of merchants and radicals within the colony. Do not provide the comprehensive evaluation until I tell you to do so.

5. Debate Practice

This is a debate simulation on the economic and environmental impacts of renewable energy policies. I will take on the role of an environmental policy advocate arguing for the transition to renewable energy and emphasizing its long-term economic benefits. Your role as the AI is to act as a spokesperson for the fossil fuel industry, defending the importance of fossil fuels for economic stability and job preservation. We will start with opening arguments, proceed with a series of rebuttals, and then deliver closing statements. During the debate, challenge my claims with counterarguments and data, pushing me to defend my stance with evidence and logical reasoning. At the end of the debate, when I tell you the debate is complete, provide a comprehensive assessment of my performance, noting the strength of my arguments, the effectiveness of my use of evidence, and suggestions for how I could improve my debating skills. Do not provide the comprehensive evaluation until I tell you to do so.

Engaging with the AI in a Simulation

Now that you’ve learned how to build a simulation, it’s time to engage with it. The real learning happens not just in creating the scenario but in your ongoing interaction with it. Engaging with the AI in simulations offers a dynamic and iterative way to apply the skills you’re learning. This iterative process—where you try out strategies, receive feedback, and refine your approach—mirrors real-world problem-solving, giving you the opportunity to learn and grow with each attempt.

However, simply running through a simulation is not enough to foster deep learning. The key lies in metacognitive reflection — your ability to critically assess your performance, the feedback you receive, and how the context of the simulation influenced both. This reflective process requires you, as the student, to think carefully about your own actions within the simulation, the AI’s responses, and how all the elements (roles, context, feedback) come together to shape the scenario. It’s about more than just going through the motions; it’s about engaging thoughtfully with the entire experience.

For instance, in the job interview example, it’s not enough to answer questions and get feedback from the AI. You need to think about how you approached each question: What strategies did you use? Did you tailor your answers to the job description? How did the AI’s feedback help you understand your strengths and areas for improvement? Reflecting on your responses, the AI’s evaluation, and how well you applied your knowledge is where the deeper learning takes place.

The value doesn’t just come from running the simulation; it comes from creating it. Designing a scenario requires you to deeply understand the roles and the context, which reinforces your knowledge of the subject matter. For example, when you build a business negotiation scenario, you must think through both sides of the negotiation, understanding not only your own objectives but also the priorities and tactics of the other party. This deepens your understanding of the theories and strategies involved, making your engagement with the material more meaningful.

Ultimately, the process of interacting with the AI simulation is not about finding a perfect answer but about practicing and refining your skills. With each iteration, you become more adept at applying your knowledge, and the act of engaging in conversations with the AI helps to solidify what you’ve learned. Through repeated engagement, you can explore new strategies, take risks in a safe environment, and adapt your approach as you learn from your experiences.

Safe and Ethical Use Note

When using these prompts, you can use what whatever GenAI tool you want. We’ve used Copilot in these examples. Microsoft CopilotLinks to an external site. is a university-supported tool that includes some data protection featuresLinks to an external site.. If you use Copilot, make sure you are logged into Microsoft 365 with your IU credentials.Links to an external site.

Using the Microsoft Edge browser will allow you to use the integrated Microsoft Copilot in the sidebarLinks to an external site. to interact with open documents such as web pages and PDFs.

Using other tools may carry with it privacy risksLinks to an external site., such as the tool using the data shared with it for future model training. As such, if you decide to use a non-IU-supported tool, you assume the related risk. Additionally, it is against IU policy to share with such third-party tools any university dataLinks to an external site.Links to an external site..

Also, note that you should not submit to any GenAI system data that may be considered another's intellectual property, unless you have explicit permission or rights to do so or if what you are sharing with the AI system is your own intellectual property.

Finally, given that the ethical guidelines for when GenAI use is acceptable varies across the contexts, including across classes, we recommend you keep a record of your interactions with the AI tool; many tools, including Copilot, keep a history of your chats, which you can reference and copy should you need to share with your instructors. Some instructors may require this.

References (click to expand)

Chernikova, O., Heitzmann, N., Stadler, M., Holzberger, D., Seidel, T., & Fischer, F. (2020). Simulation-Based Learning in Higher Education: A Meta-Analysis. Review of Educational Research, 90(4), 499-541. https://doi.org/10.3102/0034654320933544

Key Takeaways

Define roles clearly. Clearly define both your role and the AI’s role in the simulation to ensure the interaction is meaningful and aligned with the scenario’s goals.
Provide detailed context. The richer the context you provide—such as background information, relevant theories, and real-world stakes—the more effective and realistic the simulation will be. (But also make sure you are aware of the data privacy and security issues in providing such information.)
Leverage feedback for improvement. Use both immediate and comprehensive feedback to refine your decisions and actions throughout the simulation, allowing you to adjust in real-time and reflect after the scenario is complete.
Build and adapt scenarios for learning needs. The act of designing your own simulations helps you apply course concepts in meaningful, personalized ways, while also reinforcing your grasp of the subject matter.
Repeat and refine for mastery. The value of AI simulations comes not just from a single interaction but from the ability to revisit and refine the scenarios over time through iterative practice.

On the next page, we'll delve into how GenAI can act as a personalized tutor, guiding your learning through inquiry-based methods, encouraging critical thinking, and providing tailored feedback that builds your understanding step by step.