Can AI Turn Students’ Interests into Better Math Lessons?

A mathematics problem about concert tickets may sound more inviting to a music-loving student than one about an unfamiliar business. Yet changing the setting does not necessarily improve the mathematics. The quantities still need to make sense, the problem needs an appropriate level of challenge, and the context should help students think about the relationship being taught.

Generative AI makes it easier to produce different versions of a task. That creates a promising opportunity for teachers, along with a responsibility to examine what the changes actually accomplish. Personalization can involve much more than adding a child's name or a favorite sport to a worksheet.

A new study looks at teachers' decisions

A study published November 2, 2026 examined seven middle-school mathematics teachers using ChatGPT to adapt curriculum problems to students' interests. Their assignments reached 521 seventh-grade students. Researchers studied teachers' prompts, time investment, and student responses.

Teachers revised issues involving realism, depth, ethics, and cultural references. Students often wanted more specific connections to their interests than the broader versions teachers produced. The study offers an exploratory account of task design and student reactions. It does not establish that AI personalization raises achievement or saves every teacher time.

Its practical value lies in making teachers' judgment visible. A system can generate a plausible story quickly, while a teacher must still decide whether the story serves the intended learning.

Keep the mathematical relationship intact

Start with the concept before asking for an entertaining context. If students are learning proportional reasoning, the teacher should know which quantities vary together and why. A revised problem should preserve that structure rather than merely retain the same numbers.

For an illustrative example, a teacher could adapt a recipe problem into planning refreshments for a school performance. Students might calculate how quantities change when the audience doubles. The context becomes useful when it helps them explain the multiplicative relationship.

If the AI adds a fixed delivery fee, the relationship changes. That could support a different worthwhile lesson, but it would no longer be the same proportional situation. Reviewing the mathematics means checking the assumptions and working through the solution independently, including alternative interpretations a student might reasonably make.

Make the connection believable

Students can recognize an artificial reference. A task about a favorite game may feel superficial if its rules, actions, or quantities are wrong. A highly specific reference can also confuse classmates who do not share that interest.

Teachers can ask students which activities genuinely involve counting, measuring, comparing, estimating, or making decisions. Those answers may produce stronger contexts than assuming that all children of a particular age like the same things.

Offer a small number of accessible choices when practical. One student might explore a sports context, another a garden, and another a music event, while all investigate the same underlying relationship. Students should not need specialist hobby knowledge to access the mathematics.

Protect challenge and dignity

Interest and readiness are different. Enjoying basketball does not reveal how much support a student needs with ratios. A personalized story should not become an excuse to lower expectations or assign a learner permanently to a narrow category.

An advanced learner could compare two models, identify a hidden assumption, or explain when a proportional approach stops working. Another learner may need a diagram, clearer language, or a smaller first step. These decisions come from evidence of understanding, not from the theme selected for the problem.

There is also no need to upload a child's identifying details to generate a useful example. A teacher can describe a general interest and grade-level goal while following the school's approved tools and privacy arrangements. Relevant context rarely requires a detailed personal profile.

Give students a role in improving the task

After solving, students could identify one detail that made the situation realistic or one assumption they would change. This turns personalization into a conversation about models and meaning. A mathematically correct answer can coexist with an unrealistic scenario, and noticing that distinction is useful intellectual work.

A teacher might compare an original task and an adapted version across a few lessons. Student explanations, participation, and preparation time can help determine whether the extra work is worthwhile. A more attractive worksheet alone is insufficient evidence.

Begin with one worthwhile adaptation

Choose a curriculum problem that students find difficult to enter. Ask AI for a few possible contexts, check every solution, and select one that clarifies the mathematical relationship. Invite feedback and then use a new context to see whether students can apply the idea independently.

Keep the original problem beside the adaptation and ask: what has become easier to understand, and what mathematical demand must remain? AI can propose possibilities. Teachers and learners can judge whether the revised task brings the mathematics into focus.

Illustrative image generated with AI.