Sapiens Forward

Dealing with AI and trying to find happiness in this new age.

Teacher Shortages and Education Budget Gaps in the US Will Drive Adoption of AI for Education

Written by

in

The United States K-12 educational system is at a fork in the road. Recent data from the last several years shows widespread vacancies and understaffing in schools, while many districts face tight budgets amid expiring federal aid and rising costs. With the rapid increase in AI system capabilities, school districts and policymakers will inevitably look to these tools for a solution.

A Quick Summary

  • The US faces severe K-12 teacher shortages and widespread education budgeting issues.
  • AI will be increasingly adopted by school districts to augment teachers, automate tasks, and manage budget strains.
  • Full AI replacement of teachers is unlikely due to emotional, developmental, and complexity challenges.
  • But with AI being unavoidable, hybrid human-AI teaching models will offer the best alternative positive path forward.
  • Careful implementation is critical to ensure AI enhances, rather than harms, educational outcomes.

Extent of the Teacher Shortage

The teacher shortage is a nationwide issue affecting virtually every state and subject area. Recent surveys by the National Center for Education Statistics (NCES) found that nearly half of all public schools started the 2023-24 school year feeling understaffed. By the 2024-25 school year, 74% of schools reported having difficulty filling one or more teaching vacancies.

A 2024 Learning Policy Institute review estimated over 406,000 teaching positions were either vacant or held by underqualified teachers, roughly one in every eight K-12 teaching positions. About 41,000 positions were reported entirely unfilled, and an even larger number (over 365,000) were filled by educators lacking full certification for their assignments. And these shortages are not uniform, they vary by state, district, and subject.

Not Consistent Everywhere Though

Schools report that certain teaching roles are especially hard to staff. General elementary teachers and special education teachers were the most common vacancies. Other chronic shortage areas include science, math, and foreign languages. But there are differences between districts as well. High-poverty and high-minority schools are hit hardest. In 2021 to 2022, schools serving more low-income students were significantly more likely to have vacancies and staffing problems than more affluent schools. Rural districts and certain regions (often in the South and Southwest of the United States) also face acute shortages in part due to lower pay and smaller teacher pipelines.

Why Is There a Teacher Shortage?

Why there is a teacher shortage is a different sort of question, one which I’m not focused on here, but the teacher shortage probably arises from a convergence of economic disincentives and job stresses. Low pay, high stress, and fewer newcomers create a cycle that is hard to break. Specialized fields like STEM and special ed face the deepest shortages due to even greater training requirements and alternative career options for potential candidates. 

Sources:

Extent of Budgeting Challenges

American K-12 budgeting challenges are being driven by tightening revenues, rising costs, and funding gaps between jurisdictions. Nationally, per-pupil spending has reached an all-time high in nominal terms, but many districts are struggling to maintain services without the recent temporary boost of federal aid.

In the 2021-22 school year, public elementary and secondary schools in the US spent about $768 billion in total, averaging $15,591 per pupil. This marked a modest 1.8% increase in spending from the prior year after adjusting for inflation​.

This slight growth in spending was largely propped up by temporary federal COVID-19 aid. Federal funding for K-12 jumped by 32% between 2021 and 2022, pushing the federal share of education funding to 13.7%, the highest in over three decades​. This was due to relief programs under the CARES Act, and the American Rescue Plan. By contrast, state revenues for schools declined 2.6% and local revenues fell 2.0% in that period. In fact, 24 states (plus D.C.) saw per-pupil revenue decrease from 2021 to 2022. The influx of one-time federal dollars masked these declines, essentially creating a “fiscal cliff” with the expiration of federal aid and pandemic relief funds in 2024.

Much of the approximately $190 billion in federal K-12 emergency aid was probably spent or gone by the 2023-24 school year, potentially leaving budget gaps. For example, the federal K-12 budget in FY 2022 was $96 billion, down 67% from 2021 once the bulk of emergency funds had been disbursed. Many districts used relief money for tutors, technology, facility upgrades, and even hiring staff. So as these funds dried up they now face tough choices to avoid deficits.

Compounding this issue is the impact of rising inflation. Costs for salaries, transportation, food, and utilities have all climbed. While many states increased teacher pay in nominal terms in 2022 to 2023, inflation ate into those gains (e.g. the average starting teacher salary rose 3.9% in 2022 to 2023, but real starting salaries still lag 2008 levels after inflation. Districts face pressure to raise wages to attract staff (adding budget strain) even as the real value of each dollar has fallen. Essentially, budget growth has not fully kept pace with rising expenses, squeezing resources.

And finally, education funding remains highly uneven across states and districts, causing challenges with hiring and student outcomes. In 2022, per-student spending ranged from under $9,500 in Utah to over $29,000 in New York. Wealthier states and localities can spend far more, whereas high-need districts often have less funding. Many states’ funding formulas do not sufficiently account for poverty or require substantial local revenue, leading to chronic underfunding in less affluent areas. 

According to an Education Law Center analysis, some states invest thousands less per pupil in high-poverty districts compared to low-poverty districts, perpetuating inequities. These disparities are part of the budgeting challenge in that some regions face much tighter budgets than others, impacting their ability to offer competitive teacher salaries or smaller classes.

Why is There an Educational Budgeting Issue?

Again, I’m not as focused here on solving the why of this overarching question, but rather concentrated on the AI-related side-effects. But to summarize, the pervasive K-12 budgeting issues in the US largely appear to arise from mismatches between resources and needs. Temporary fixes (like federal aid) are waning, while long-term issues (like inequitable funding and policy-driven resource shifts) intensify. Without adjustments, many districts will have to make difficult cuts or seek new revenue to balance their budgets.

Sources:

Integration of AI in Education Will Become the Solution to Teacher Shortages and Education Budget Gaps in the US

AI Will be Pursued as a Solution to Budget and Hiring Challenges

One of the primary reasons AI will be pursued for education is the potential for substantial cost savings compared to maintaining or expanding human teacher workforces. AI systems, once developed and deployed, should be expected to have relatively low marginal costs, especially when scaled across many classrooms or districts. In contrast, hiring additional human teachers involves ongoing salary expenses, healthcare benefits, retirement contributions, and professional development costs, all of which compound annually with inflation. AI platforms can help automate or assist with routine tasks such as grading, administrative scheduling, and even tutoring support, reducing the need for additional paraprofessionals and administrative staff. Furthermore, AI services can often be provided on a subscription or licensing model that, while requiring upfront investment, offers predictable and often lower long-term operating costs than expanding human staffing. 

These economic considerations will likely drive districts facing tight budgets and teacher shortages to adopt AI solutions as a pragmatic cost-control measure, even in cases where student outcomes will be in jeopardy.

AI as Augmentation Support for Teachers

The current reality is many schools are already experimenting with AI to support, not replace, teachers. AI is currently being used to generate lesson plans through platforms like Diffit, Curipod, and MagicSchool.ai. AI tools are also being explored to assist with grading and feedback using solutions like EssayGrader and Gradescope. In addition, AI can potentially be used to identify student learning gaps through adaptive learning software and AI-driven formative assessments. AI tutoring bots can provide individual tutoring support after class hours, while administrative tasks like scheduling and communication with parents are increasingly being handled by AI solutions.

AI’s impact on the teacher shortage and budgets could be significant. By handling repetitive tasks, AI allows teachers to focus more on instruction, increasing their productivity and potentially reducing burnout, which could lead to improved retention. Additionally, the need for paraprofessionals or extra staff to assist with administrative and tutoring roles may decrease, resulting in cost savings in the area of staff budgeting. Furthermore, AI supports differentiated learning by enabling teachers to personalize instruction more easily for each student.

The idea is that when deployed at scale, the AI tools will have a lower cost (compared to staffing with human teachers) and will be more consistent to deploy to districts with hiring challenges. The fact that there is already a teacher shortage suggests that this kind of implementation will not negatively impact existing teachers’ jobs or employment.

AI is not a full replacement for teachers’ human judgment, emotional support, or classroom management capabilities. Teachers must undergo training to work effectively with AI, a process that requires both time and financial resources. Additionally, there are equity concerns, as wealthier districts may adopt AI tools more quickly, potentially widening the digital divide between affluent and less affluent schools.

The reality is AI augmentation is already happening. It could have potential to make teachers’ jobs easier, improve student outcomes, and stretch limited budgets, but that is an incredibly optimistic leaning point of view. This is a very new and dramatic shift to the field of education. It is likely that effective solutions are going to be identified through trial and error, with vast amounts of young students and teachers being test subjects in an active teaching experiment on a grand scale.

Will AI Replace Human Teachers?

I don’t believe teachers should be replaced by AI, I think that would have negative implications on the social and intellectual development of young people. Thankfully, serious challenges emerge when considering “Teacher Replacement AI” as a solution.

One preventative issue is the social-emotional learning gap, as students need human interaction to develop social, emotional, and collaborative skills that are crucial for future workplaces and life. AI cannot yet replicate deep emotional intelligence, mentorship, or the creation of a vibrant classroom culture. Equity and access problems also arise, with students lacking reliable internet or devices being left further behind. Moreover, students requiring extra help, particularly those with disabilities or who are English learners, often need personalized human support that AI struggles to fully provide.

Secondly, overall teaching process complexity introduces a preventative challenge to teacher replacement. Teaching involves more than delivering information; it requires managing group dynamics, motivating disengaged students, handling behavior problems, and adjusting instruction based on subtle, real-time feedback, areas where skilled human teachers excel and AI still struggles. Public resistance and ethical concerns are significant as well, with parents, unions, and many educators opposing the replacement of teachers by machines. Child development experts emphasize the risks of screen time overload and highlight the need for relational, in-person learning.

Finally, there is a quality risk, as fully AI-driven education could devolve into a “factory model” of learning with poor engagement. Students, especially young children, arguably thrive best through play, exploration, conversation, and relationships, not through passive information reception.

So, in my opinion using AI to completely replace teachers is not yet currently viable, and even if technological capabilities were to advance further, such a shift would pose serious educational, ethical, and social risks that every reasonable parent and educator would want to avoid – if possible. 

However, another unfortunate reality though is that teacher shortages and education budget shortfalls don’t impact areas in a consistent and equitable manner. Sadly, demographically disadvantaged or lower-income districts face sharper issues with staffing and budgets, and those challenges will likely lead those areas to rely upon more drastic and aggressive solutions involving AI teaching tools. Districts in that kind of situation might be more likely to explore Teacher Replacement AI, even if it is demonstrably proven that student outcomes are less optimal.

How Far Could Educational AI Advance in 5 Years?

Rapid AI evolution is likely over the next decade and in education that could mean even more advanced AI tutors and teaching aids. I imagine there could be AI education systems that simulate and respond to rich dialogue, provide more personalized learning paths, and AI teaching assistants embedded in classrooms that help teachers monitor student engagement in real-time. But I still remain skeptical, and even in 5 years I don’t believe AI will completely replace teachers, except in maybe poorer districts or extremely remote areas that might be tempted to use AI to replace teachers very aggressively out of desperation.

Why I Believe Full Teacher Replacement AI Will Remain Unlikely

While AI may improve at detecting emotions, building trust, nurturing resilience, and mentoring are deeply human roles, critical especially in early childhood and adolescence. Younger students require socialization, teamwork, conflict resolution, empathy development, and human role models, none of which AI can fully deliver. Additionally, real learning is messy, emotional, relational, and personal. No AI, however sophisticated, can replicate the human-to-human “aha!” moments, inspiration, or guidance at critical times.

When it comes down to educational policy, I feel the majority of our society places a high value on human educators, particularly in public K-12 education, even if rhetoric about teacher pay and unions in the US can often be politicized. Ultimately, I believe teacher unions, parents, and many policymakers would strongly oppose full automation.

The Best Chance for a Positive Outcome System is Probably a Hybrid Human-AI Teaching Model

Acknowledging that AI is here to stay, a hybrid model seems the most plausible path to maximize benefits and minimize downsides. In such a model, one teacher could manage a larger class more effectively because AI would help personalize content for each student. Teachers could specialize more, focusing on coaching, mentoring, and fostering deep thinking, while AI handles repetitive and scalable tasks. Additionally, more students could access “individualized” attention through AI-powered tutoring, with the human teacher remaining the central anchor for guidance and support. Rather than being replaced, future teachers could be empowered, amplified, and elevated through partnerships with AI.

If we take one of my AI Principles, and accept that AI is here now and there isn’t much we can do to “put that genie back in the bottle”, then we must strive to figure out where we can find positives to emphasize.

Final Thoughts

If investments are made thoughtfully, AI could help solve part of the teacher shortage and funding challenges, not by eliminating teachers, but by making teaching more sustainable, rewarding, and scalable. But there will be big risks with implementing AI in education. We must ensure that a human connection remains at the heart of learning. And we have to strive to prevent AI-only education models from proliferating and inadvertently creating even more negative challenges scenarios for students in disadvantaged school districts.

I remain hopeful that we will take the path that most benefits all of us, but it will take involvement and diligence by everyone to make that hope realized.

[mailerlite_form form_id=3]

More posts