Key points:
- Institutions must guide students to use AI thoughtfully and ethically
- Leveraging AI for institutional improvement, continuous learning
- Exploring generative AI through the lens of science fiction: A framework for educational dialogue
- For more news on ethical AI, visit eCN’s AI in Education hub
AI is rapidly reshaping academia, creating both opportunities and challenges for institutions. As generative AI tools gain traction in workplaces, universities have a responsibility to prepare students for an AI-driven future. Restricting AI or moving back to solely pen-and-paper assessments to avoid cheating isn’t just impractical–it’s counterproductive. The proverbial horses have fled the barn, and there’s no going back.
Instead, institutions need to shift from a “prevention” mindset to ethical integration. Take calculators, for example. When they were first introduced, many professionals resisted them, fearing they would weaken students’ math skills. Today, calculators are indispensable learning tools in classrooms, with apps on nearly every device we own. Generative AI is on a similar trajectory. It’s time for educators to recognize it as a valuable asset, rather than a threat or shortcut that undermines learning, so students benefit from its advantages while also developing critical skills.
Reshaping perspectives
To integrate AI effectively, universities need to shift their perspective: AI isn’t a substitute for learning–it’s a tool to deepen their understanding. Generative AI, at its core, is a language model, not a knowledge model, so students must learn to verify AI outputs to ensure accuracy, rather than relying on it for quick solutions and complete answers.
When used responsibly, AI can guide students through the problem-solving process, teaching them how to approach complex questions independently. For this shift to happen, curriculums must evolve to teach students to not use Gen AI as a crutch, but rather as a guide for the problem-solving process. One of my research areas is Intelligent Tutoring Systems (ITS), which has been exploring ways to prompt Gen AI sites to offer questions, not answers. This approach fosters computational thinking while incorporating Gen AI into their learning process.
To make this shift, curricula must evolve. Educators should design lessons that incorporate AI as part of the learning process, showing students how to use these tools ethically and responsibly. This doesn’t mean abandoning traditional methods entirely–it means adapting them to reflect the tools and technologies students will encounter in their professional lives.
Rethinking assessment design
Even before Gen AI, it had become increasingly easy for students to access assignment answers online. This raised long-standing questions on how to encourage genuine effort and accurately assess understanding. Some institutions have reverted to pen-and-paper tasks, but this is a disservice to students who will use technology in their careers.
For example, as AI use in coding grows, students still need coding fundamentals to refine and verify AI outputs. Coding curriculums should shift, incorporating AI to supplement learning rather than supplanting it.
In my courses, I’ve been customizing assignments to require original thinking, preventing students from simply copying AI answers. In an advanced computer science course, my students have to implement a version of the Yahtzee game this year. Because thousands of solutions are readily available online, I modified the assignment’s specifications to ensure students won’t find exact pre-written code online. Even if they source their code from the web, they must understand and modify it to meet my criteria, teaching them valuable collaboration skills.
This philosophy can be applied across disciplines. By designing assignments that require creative thinking, customization, and critical analysis, educators can ensure students engage deeply with the material, even when using AI as a tool.
Reforming institutional structures
Successfully integrating AI into academia requires more than updated assignments and lesson plans–it calls for systemic change. Universities must ensure equitable access to AI tools so that all students, regardless of their economic background, can benefit.
Ensuring equal access to AI tools can help bridge gaps between economically disadvantaged students and those with more resources. For example, schools could offer blanket subscriptions to commonly used AI tools, much like they have done with laptops or tablets.
Equally important is the development of standardized curricula that teach AI skills ethically and effectively. For example, CS2023–a set of curriculum standards developed by the Association for Computing Machinery (ACM), IEEE Computer Society, and the Association for the Advancement of Artificial Intelligence–offers clear guidance on the knowledge and skills students should gain in computer science and related disciplines at the undergraduate level in an AI-driven world.
In addition to curriculum reform, institutions need to establish security practices to prevent misuse like unauthorized access to school systems or academic dishonesty. Clear guidelines on the appropriate use of AI can help maintain academic integrity while allowing students to benefit from these powerful tools.
Preparing for the AI-driven future
The integration of AI in higher education isn’t optional; it’s essential. Educational institutions must embrace AI integration while maintaining the academic mission by ensuring authentic learning, developing innovative assessment methods, and providing equitable access.
As AI becomes increasingly central to industries ranging from healthcare to finance to creative arts, the role of educators is clear. It’s not just about teaching students to use technology–it’s about guiding them to leverage it thoughtfully and ethically. Universities that rise to this challenge will empower their students to enter the workforce not only as capable professionals but as confident, responsible innovators.
Dr. Amruth N. Kumar is Professor of Computer Science at Ramapo College of New Jersey and is ACM Chair of the ACM/IEEE-CS/AAAI CS2023 Steering Committee.