Schools and education organizations have poured resources into artificial intelligence, yet many struggle to show clear financial gains from these investments. IDC's 2026 Worldwide AI Economics and Cost of Ownership Survey found that 69% of organizations see a sizable gap between what they spend on AI and the financial results they can actually track. Only 15% of those surveyed reported their AI return on investment topped 100%.
Financial Pressures and AI Adoption in Education
Those rolling out AI in education face steep costs. In a separate IDC survey from May 2026, 73% of respondents flagged high AI expenses as a major risk. These numbers point to the ongoing difficulty of justifying AI budgets when benefits are hard to spot or measure.
The State Educational Technology Directors Association (SETDA) reports that artificial intelligence remains the top tech focus for state education leaders in the U.S. In October 2026, SETDA said 79% of surveyed state chief information officers, education-technology directors, and agency leaders across 45 states and one territory named AI as a leading priority, up from 68% in 2025. Eighty-three percent of states said they had an AI initiative. Still, 52% of respondents said their states collect no data on how education-technology products are used or how effective they are, showing a major gap between AI adoption and the ability to prove results (GovTech).
SETDA's 2026 data also shows a move from broad ed-tech rollouts toward tighter accountability and clearer purchasing rules. Fifty-three percent of states had published guidance for buying AI-enabled tools, and 79% of those included evidence standards in that guidance, up from 45% in 2025. This shift signals that institutions want proof of results and solid vendor claims before spending more.
In higher education, a 2026 survey summarized by the American Association of Collegiate Registrars and Admissions Officers found that about 70% of provosts used AI at least weekly. Yet only around 10% said their institution had a central AI strategy. Most AI use happens through scattered experiments, not coordinated planning, and few provosts saw AI as changing how their institutions actually operate (AACRAO).
Roy Y. Chan's 2026 review of the 100 universities in the Times Higher Education rankings found most schools sit in the middle of the AI development curve. AI is often limited to research centers, degree programs, and policy statements, rather than being part of daily governance and operations. Only four universities had fully integrated AI across the institution, showing how rare true enterprise-wide AI adoption remains in higher education.
Decision-Making and Integration Barriers
Funding decisions for AI in education now draw in many stakeholders. IDC's September 2026 C-Suite Tech Survey found that chief information officers and chief technology officers took part in 63% of AI funding decisions, business-function executives in 62%, and chief financial officers in 61%. Another IDC survey found that 42% of CEOs planned to hire a chief AI officer within the next year.
Even as interest grows, buyers keep raising concerns about AI vendors' products and contract terms. Limited integration with current systems, lack of transparency in models and data, and worries about vendor lock-in each came up for 34% of respondents. Overpromising and underdelivering, along with unclear pricing, were each cited by 31%. These sticking points can slow AI adoption in education and make it harder for institutions to get full value from their investments.
Across the sector, schools and colleges are pushing AI suppliers to show value for money and real returns. In October 2026, Jisc, Universities UK, GuildHE, and the Association of Colleges issued a joint statement with six expectations for AI suppliers: transparency at contract signing, protection of value over time, usage visibility and cost control, supply-chain accountability, and continuity when large-language models change. The statement pointed out that unclear pricing and usage models make it tough to predict costs or prove value for money.
Policy, Accountability, and Evidence Standards
Policy changes are shaping how AI spending plays out in education. In July 2025, the U.S. Department of Education issued guidance stressing responsible use, privacy protections, and stakeholder involvement for AI projects funded by federal education grants. Institutions must make sure AI purchases fit grant goals and legal duties.
State laws are also affecting AI adoption. North Carolina's 2026 education budget set aside $500,000 for schools to buy Khan Academy's Khanmigo licenses for students in grades 6 through 12. This was cut from a proposed $10 million after educators and AI experts questioned the program's evaluation, showing how evidence and review concerns can directly shape public AI funding.
Indiana's Consumer Data Protection Act, effective January 1, 2026, brings in data-protection rules that matter for AI-enabled education services. These rules affect vendor choices, contracts, and risk reviews for institutions. Compliance adds indirect costs and governance steps beyond just buying AI software.
Stanford's 2026 research agenda on AI and learning, written by C. D. Robinson, C. Agnew, and Susanna Loeb, calls for stronger evidence to back claims of real K–12 transformation. The agenda stresses the need for research designs, outcome measures, and proof of implementation to show whether AI tools actually improve learning. This highlights why reliable evidence of educational effectiveness is key for telling apart productive investment from tech adoption that lacks proven benefit.
Looking Ahead for AI in Education
IDC projects that AI infrastructure capacity could meet demand by 2029, which may help ease current cost pressures. Until then, education organizations and other enterprise buyers will keep pressing vendors for solid proof of long-term financial value, strong security, and sound governance. IDC's analysis offers more detail on what customers expect from AI technology vendors.
