Engineering judgment in an AI-assisted world

This is a story about flying airplanes. (This is actually a story about AI.)

When I was training for my private pilot’s license a few years ago, I was surprised by how much I had to learn that had seemingly nothing to do with actually flying the plane. To earn a license, you must demonstrate knowledge of aircraft systems, weather, regulations and physics, among other things. The point of all this is not to become an aircraft mechanic, a meteorologist or an aerospace engineer; rather, this knowledge is necessary to be a safe and effective pilot. A pilot relies on sophisticated systems, but cannot do so without also exercising human judgment. When something unexpected happens, that knowledge allows a pilot to troubleshoot, recover and complete the flight safely.

I think about that training often now, especially when prospective students and parents ask how the College of Engineering is preparing engineers for an AI-enabled professional world.

One college strategic priority is to become an AI-native college of engineering. As an educator, I understand this to be much more than simply giving students access to new tools and technologies. Successful engineers have always needed to know how to select the most effective tools to solve problems, and AI adds a powerful new option (along with a new set of responsibilities) to that toolset.

Becoming AI-native means being prepared to make informed decisions about when AI improves an engineer’s work, when it does not, and how using AI changes the responsibilities of the engineer. Just as flying an aircraft is more than the ability to manipulate controls, using AI as an engineer is more than the ability to enter a prompt. Engineers need to understand AI’s technical foundations, its limitations and the ethical and environmental considerations of using it. To me, AI fluency is not about being able to use AI everywhere—it is the ability to make sound judgments about when and why to use it.

Toward that end, the College of Engineering wants every engineering student to develop a common foundation in AI and then learn to apply it within their own field, incorporating experiential learning and ethics throughout. In fall 2026, the college launched the undergraduate Certificate in Applications of AI in Engineering, which offers students one pathway toward that common foundation.

Beyond that foundation, the applications of AI are discipline-specific; we are working to distinguish those disciplinary needs, rather than creating a generic curriculum for all of engineering. Our faculty are identifying what students in each major should understand about AI and how they should practice using it. For example, the Department of Chemical and Biological Engineering recently worked with the college’s Center for Innovation in Engineering Education to host a series of faculty-led workshops on AI, machine learning and education. These workshops ranged from policy discussions to technical training on large language model capabilities and limitations. Those conversations—and others like it—give departments a shared understanding of what AI-native graduates in their disciplines should know and be able to do.

These educational initiatives extend throughout our curriculum. For example, in a course on professional ethics for master’s students in STEM, Laura Grossenbacher, the college’s director of technical communication, has developed assignments that ask students to use multiple large language models to analyze ethical case studies and present their findings. Students must answer questions such as whether AI tools agree, whether their conclusions are defensible and what they missed. The most interesting situation comes when AI gives a seemingly plausible but incomplete answer (something I have encountered many times in my own teaching). In these cases, students are asked to identify what is missing and defend their reasoning rather than simply deferring to AI.

That judgment is one of the core skills we want future engineers to develop. At all levels of engineering education, our goal remains consistent: We are not measuring AI fluency by students’ ability to produce an AI-generated answer, but by their ability to evaluate that answer, explain what it means and decide what to do next.

As I begin serving as director of the Center for Innovation in Engineering Education, one of my priorities is to create a living map of the innovative educational practices related to AI already underway in our college. I am still new in this role, and my first responsibility is to listen in order to understand what faculty are already trying and where our center can genuinely help. Faculty should be able to find colleagues who have tried something similar, locate an example or case study and build on what others have learned, rather than starting from scratch.

This map will also help center staff identify common challenges, direct support where it is most useful and help faculty adopt and adapt promising practices for new courses and student populations. Through its “Engineering 4 All” Pathways program, the center already brings faculty together in learning communities where they can explore and test ideas. The center’s next step will be to turn what these communities learn into case studies and course materials that others can use. This is the role I envision for the Center for Innovation in Engineering Education: connecting educational innovators, equipping them with the tools they need and amplifying their successes.

My training as a pilot taught me that sophisticated tools are most useful when the person operating them understands both their capabilities and their limitations. I want our students to bring that same critical judgment to AI. The tools we use will continue to change, but our responsibility as educators will not.

We must prepare future-ready engineers who can evaluate what tools to use, when to use them, and why. AI is the immediate challenge, but the lasting impact to engineering education is to build a community of educators who learn from each other, evaluate what works and make successful ideas easier to adopt. This is how we ensure we are ready for the next technological change—not just this one.

Amanda Smith directs the College of Engineering Center for Innovation in Engineering Education at UW-Madison. Her work focuses on advancing engineering education through curriculum innovation, faculty development, and creating programs and partnerships that help students succeed.