Generative AI (GenAI) has become one of the most talked-about technologies in higher education. From drafting reports to writing code, students and faculty are exploring how tools like ChatGPT, Claude, Gemini, and Copilot can support their work. But what does AI mean for those learning and teaching mathematical optimization?
Earlier this year, we published our inaugural State of Mathematical Optimization in Academia Report, based on our survey of students and faculty currently using Gurobi.
We asked them a variety of questions about how they’re applying mathematical optimization in the classroom—from the types of problems they’re solving to their preferred programming languages.
But in today’s age of AI, we’d be remiss not to ask them how GenAI is impacting their work, and how they see optimization and GenAI working together. Below are highlights from the report.
AI Is Already Influencing Academic Research
One of the survey's clearest findings is how much GenAI has become part of academic workflows.
Among respondents, 85% of students and 82% of faculty reported that GenAI has a “moderate to significant” impact on their research. Respondents also reported similar impacts on report writing, teaching (for faculty), and class assignments (for students).
These findings highlight an important reality: many students are already incorporating AI tools into their learning process. Rather than viewing AI as a separate topic, educators may increasingly need to consider how these tools fit within optimization curricula and classroom expectations.
Coding Leads the Way
We also asked our academic community how they think GenAI can help with optimization.
Coding emerged as the top use case by a wide margin, with 80% of faculty and 79% of students saying that GenAI will help with coding tasks.
Faculty respondents also cited model generation (55%) as a top use case for combining GenAI with optimization, with students sharing a similar perspective (53%).
Specialized AI agents, such as Gurobi’s Modeler, are already available as companions during the modeling process itself—helping users move from problem descriptions to building optimization models more quickly.
Lowering the Learning Curve
Of course, optimization still requires human judgment and oversight. AI can assist with coding and formulation, but practitioners must still determine whether a model accurately represents their real-world problem. By reducing some of the friction involved in implementation, GenAI can make optimization more accessible and allow learners to spend more time focusing on the underlying optimization concepts.
“GenAI can serve as an initial tutor and help students understand how to apply solutions or overcome modeling challenges,” says Dr. Laura Albert, Professor of Industrial and Systems Engineering and Senior Associate Dean for Faculty Affairs and Development at UW Madison’s College of Engineering.
Several respondents of our survey described using AI tools to learn new modeling approaches, explore solver features, and discover alternative ways to implement optimization models.
In addition, nearly a quarter of students reported using Gurobot, Gurobi's AI-powered optimization assistant, and 60% of those who used it found it "very helpful." Faculty usage was lower but still notable, with one in five faculty members reporting that they had used the tool. Among those users, 79% reported satisfaction with its responses.
Such tools may be particularly valuable for students who are learning independently, exploring advanced topics, or working outside traditional operations research programs.
Not a Replacement for Skill Development
While GenAI can be a powerful learning assistant, many in the academic community have emphasized that it should complement learning and help break down barriers to entry.
As GenAI becomes more deeply integrated into academic workflows, students will still need to develop the modeling, analytical, and critical-thinking skills required to formulate optimization problems, evaluate results, and determine whether a solution truly addresses the challenge at hand.
The Future Is Bigger Than GenAI
The growing adoption of GenAI should not obscure an important reality: AI is much larger than large language models.
Optimization remains a foundational AI technology for decision-making, resource allocation, planning, and operational excellence. While GenAI excels at creating content and assisting users, optimization provides the mathematical rigor needed to determine the best course of action and deliver explainable, reliable decisions.
Importantly, students must still develop modeling skills, rather than mindlessly relying on AI tools without the ability to verify or refine.
While GenAI promises both more opportunities and responsibilities for academics, Dr. Albert remains optimistic about the future.
“The question is, are we going to meet the moment and democratize optimization by increasing its awareness, and expand its use? This is an exciting time for the field. It’s going to take all of us, and I can’t wait to see what happens next.”
To explore the full State of Mathematical Optimization in Academia Report, download your free copy here.
