Whether you license, build, or mix the two, the contract and the pilot decide who controls student data and who pays when usage grows.
Artificial intelligence tools are arriving on campuses faster than procurement offices can write policy. A 2025 EDUCAUSE study, cited in an October 13 Community College Daily commentary by Arapahoe Community College’s Jeff Nesheim, found only 2 percent of respondents were covering new AI costs with new funding. Most colleges are paying by cutting somewhere else. That makes every contract a budget decision as well as a technology one.
The cheapest AI contract is the one you can leave.
1. What Student Data Does It Touch?
Start with the data. Advising questions, financial aid details, and records are sensitive. Ask where data is stored, who can see it, and whether it can be used to train a vendor’s models. A chief information security officer should review the answer before anyone signs.
2. How Will Costs Grow?
Nesheim reports that software-as-a-service costs rose an average of 34 percent in the first half of 2026 and cites California State University’s $17 million annual contract for 22 campuses. Whatever the exact figures, the lesson is to ask how price changes with students, usage, and time. Seek caps on annual increases and clear definitions of what counts as usage.
3. Can We Leave?
Exit rights are easy to overlook. Ask whether the college can export its prompts, records, and configurations, and what happens to data after termination. Contracts that make leaving costly turn a pilot into a permanent commitment.
A pilot that cannot end is not a pilot.
4. Who Will Run It?
Nesheim argues colleges should consider running open-weight models on hardware they control, and his estimates suggest large savings. His own piece acknowledges that owned infrastructure needs capital, technical skill, and sustained commitment, and gives little detail on staffing. Whatever the path, name an owner, budget for training, and plan for turnover.
5. How Will We Know It Works?
Define success before the pilot starts: accuracy, staff time saved, student satisfaction, and total cost including people. Run a commercial tool and a local option side by side on one internal use case for a semester, such as helping advisors search policy documents.
6. Who Is at the Table?
Decisions touch the president, the chief information officer, the chief information security officer, the chief financial officer, and the vice presidents of academic affairs and student services. Institutional research directors often know best what data is reliable. A decision made by technology alone will miss academic and student concerns, and one made without technology will miss risk.
7. What Does It Mean for Equity?
Community colleges serve working, first-generation, and part-time students, many of whom cannot afford premium tools. If AI access depends on a vendor contract, the least funded institutions offer the least. Ask how every enrolled student will get access, not only those in a pilot.
Negotiate Before You Need To
Leverage is highest before signing. Colleges can ask for price caps, data limits, audit rights, and the ability to pause usage if outcomes disappoint. Consortia and state systems can negotiate together, which helps smaller colleges that lack the staff to review every term. Even a one-page checklist, shared across the cabinet, keeps the conversation focused on the same questions each time.
Keep a record of what was promised during sales conversations and make sure it appears in the contract. Verbal assurances about privacy or pricing rarely survive a change of account manager.
Finally, set a review date. AI products change monthly, and a contract that made sense in October may not in June. Schedule a cabinet review at the end of each term to revisit usage, cost, complaints, and alternatives, and be ready to adjust.
A Word of Caution
The commentary is advocacy from a college piloting its own approach. The cost figures are estimates, and the claim that open models match commercial quality is not benchmarked. Treat it as a reason to ask questions, not to skip them.
For the full cost claims, the caveats, and a pilot plan, read College Leads’ analysis of the case for owning AI. If you work with presidents, chief information officers, or finance leaders and want to reach them, you can build a targeted higher education contact list here.
About the Author: Charlie Isham is the Founder and CEO of K12 Data, Inc., which operates K12 Data, College Leads, and K12 Talent. He has spent more than 15 years building education data infrastructure and has watched how schools, colleges, and the companies that serve them buy for over a decade. He is based in Encinitas, California.
