Online: Trending Now

There is a new “cliff” in American higher education, and it is not the demographic cliff. Rather, it is the dramatic cliff in math knowledge, skills, and abilities.

Let me be clear that other discipline deficiencies are found in this new generation of college students, however they are dwarfed by those in math. These have most recently been quantified in a report from the University of California San Diego. The official “Senate-Administration Workgroup on Admissions Final Report” (released November 6, 2025) contains disturbing findings. This widely discussed report revealed that nearly one in eight incoming freshmen couldn’t meet middle school math standards!

Key Findings of the UCSD Report

Significant Increase in Remediation Needs: Between 2020 and 2025, the number of freshmen at UCSD with math skills below a middle-school level increased nearly thirtyfold. Roughly one in eight incoming students now requires remedial education in math.

High School Grades Are Not Correlated with Proficiency: The report found that many students with strong high school math GPAs (e.g., a 4.0 average) still placed into remedial courses designed for middle school-level material, indicating that grades are no longer a reliable measure of actual skill.

Widespread Problem: The issue is not isolated to one institution. Reports from other universities, such as George Mason University, also describe college students struggling with basic math concepts, even those in pre-calculus and calculus courses.

Systemic Issues: The decline is part of a longer-term trend exacerbated by the pandemic and is prompting calls for greater accountability in the K-12 education system and for a re-evaluation of admissions standards, such as developing a “Math Index” to better assess preparedness.

Impact on Higher Education: The increased need for remedial courses strains university resources and risks “setting them up for failure”. Higher education institutions are having to create new, basic courses to address fundamental knowledge gaps that in previous years were assumed to be covered in high school.

This is not just at UCSD or California, in an Associated Report carried by CNN titled “US high school students lose ground in math and reading, continuing yearslong decline,” there is evidence that the decline may be originally seated in the COVID-19 pandemic which forced students out of their schools and into homes and protected day care centers.

A decade-long slide in high schoolers’ reading and math performance persisted during the COVID-19 pandemic, with 12th graders’ scores dropping to their lowest level in more than 20 years, according to results released Tuesday from an exam known as the nation’s report card. Eighth-grade students also lost significant ground in science skills, according to the results from the National Assessment of Education Progress. The assessments were the first since the pandemic for eighth graders in science and 12th graders in reading and math. They reflect a downward drift across grade levels and subject areas in previous releases from NAEP, which is considered one of the best gauges of the academic progress of U.S. schools. ‘Scores for our lowest-performing students are at historic lows,’ said Matthew Soldner, the acting commissioner of the National Center for Education Statistics. ‘These results should galvanize all of us to take concerted and focused action to accelerate student learning.’

It’s not just introductory discipline-areas that are impacted. The deficits also result in deficiencies in logic, problem-solving strategy, and confidence in working with data and statistics. These are essential skills for further learning upon which we scaffold more advanced curricula. They are foundational for business, engineering, and the entire curriculum. Upon graduation, they become essential workplace skills and abilities. So, how can we solve this problem?

In years gone by, we would probably roll out dozens of sections of remedial classes that for many entering students would become prerequisites for more advanced classes. That would be expensive, not only in faculty and staff salaries, but adding remedial courses can delay progress, incurring additional expense in the completion of the learners’ degree or certificate completion. AI could provide a more personalized approach to identifying and remediating the specific skills needed. An adaptive learning approach can customize an asynchronous, modular, learning pathway that includes on-going assessments to ensure mastery of the identified knowledge and skills. It can selectively identify the specific areas of individual learner deficiency and target instruction and personalized tutoring in those areas.

In a column last August, “AI Can Facilitate Mastery Learning” I outlined the benefits of AI facilitated mastery learning. Modules can be completed at the learner’s preferred pace and pedagogical approach; adjusting to the individual’s learning preferences and time available. Modules include frequent assessments in the form of brief quizzes to ensure the learner fully understands the principles of each operation prior to moving forward to the next topic. A learner may not move to the next topic until a 90% score is achieved on the prior topic.

Included in the August column is a description of, and link to, “The Precision Learning Companion,” a prompt from “There’s an AI for That” (TAAFT) which can be used with the leading generative AI apps such as Gemini, ChatGPT, and Claude. The prompt describes itself as:

This prompt turns AI into an ultra-detailed, dynamic personal tutor that doesn’t just quiz, it teaches deeply, layer by layer, until the user genuinely masters the material. It’s built to adapt in real time, constantly diagnosing knowledge gaps, and never moving forward without full comprehension. Every answer, right or wrong, triggers a structured, narrative-style breakdown explaining the what, why, how, and broader context, ensuring true understanding.

The AI is designed to feel like having a supportive but meticulous mentor who scaffolds learning: progressively challenging the user if they perform well, slowing down and simplifying if they struggle, and always reinforcing psychological safety through encouragement. It uses textually described visual aids, memory tricks, real-world examples, and step-by-step remediation when needed. Mastery, not speed, is the goal.

Just two weeks ago, TAAFT released another prompt “Multimode Learning System,” that offers additional options for AI-driven interactive learning of specified skills and information. This offers a Navigator Mode, Tutor Mode, and Roadmap Mode to enable responsive approaches to meet the learner’s needs.

I encourage readers to test out these free prompts. You simply copy the prompt and paste it into the prompt box on one of the major generative AI apps. I believe these can be effective tools for the individual instructor who can utilize them to help bring their students up to speed on math and other skills. These tools demonstrate some of the current capabilities that could be implemented by an individual or a department as soon as next semester.

This may provide an institution’s beginning of the use of AI to remediate knowledge and skill deficits on a wide scale. Are you using these technologies to provide individualized, customized online learning to your students at their own pace for their own needs? How might this best be shared across your university to meet the needs of the emerging math cliff and other individual knowledge and skill deficits?

 

This column was originally published in Inside Higher Ed.