From AI to Infrastructure: experience: LIAISON 2026 Revealed the Next Phase of Enrollment Technology

One of the more useful ideas in innovation research is that technologies rarely mature in a straight line. They tend to follow what strategists call an S-curve. Early progress feels slow as organizations experiment with new capabilities and search for practical applications, and then, eventually, the curve steepens. Adoption accelerates, expectations shift, and attention moves from what the technology can do to how work itself should change.

Near the top of the curve, something curious happens. The technology becomes difficult to notice because it has become part of the infrastructure. Few organizations market themselves on their use of electricity, cloud computing, or relational databases. Those technologies are no longer products to be admired; they are simply assumed.

Artificial intelligence may be approaching a similar point within enrollment management.

The AI-focused sessions at this year’s experience: LIAISON conference in Baltimore offered an opportunity to consider that possibility. Rather than concentrating on generative AI as a collection of novel features, speakers and attendees explored predictive analytics, workflow automation, admissions review, communications planning, search, and simulation. Each session addressed a different operational challenge, yet together they described intelligence becoming embedded throughout the enrollment lifecycle.

The pattern becomes clearer when viewed through the S-curve. Early experiments with AI naturally focused on visible capabilities. Could a language model draft an email? Summarize meeting notes? Generate marketing copy? Those demonstrations were valuable because they established that the technology could perform tasks once thought uniquely human.

As organizations gained experience, however, different questions began to emerge. Which decisions should AI inform? Which workflows should it support? Where does additional intelligence improve institutional performance without adding unnecessary complexity? Many of the conversations in Baltimore reflected those questions.

Several sessions examined predictive and prescriptive analytics through Liaison Othot, including Predictive Insights: Data-Driven Enrollment Decisions in Action, Predictive and Prescriptive Modeling for Top-of-Funnel Activities, and Designing Data That Drives Yield Predictions and Recruiting Effectiveness. Forecasting has always been part of enrollment management, but predictive modeling changes its role. Analysis moves closer to the moment when institutions can still alter recruitment strategies, adjust communications, or redirect resources before outcomes become fixed.

Another group of sessions focused on workflow. AgentForce in Action, AI-Driven Student Outreach, and enhancements to TargetX Communication Planner explored the role of AI within recurring enrollment workflows. Communications planning, outreach, prioritization, and execution become increasingly coordinated as routine administrative work shifts toward intelligent automation, theoretically leaving admissions professionals with greater capacity for advising, relationship-building, and institutional decision-making.

Discovery appeared as another recurring theme. Prospective students increasingly encounter institutions through AI-generated responses, conversational search, recommendation engines, and synthesized information environments rather than traditional keyword searches alone. Visibility depends not only on what an institution publishes, but on whether its expertise can be interpreted, connected, and surfaced within these emerging systems.

Admissions review provided another perspective on the same evolution. Sessions exploring holistic review and AI-assisted simulation focused less on replacing professional judgment than on expanding the information available to support it. Pattern recognition, scenario modeling, and contextual insights strengthen decisions without removing the responsibility for making them.

The conference sessions ranged from predictive modeling and communications planning to search, admissions review, and workflow automation, spanning nearly every stage of the enrollment lifecycle. Intelligence appeared throughout those discussions, although rarely as the primary subject. Instead, it surfaced as an attribute of systems that help institutions forecast demand, coordinate communications, improve discovery, and support complex admissions decisions. As those capabilities become increasingly connected, the technology itself begins to recede into the background while institutional capability becomes more visible.

Organizations, meanwhile, are investing in prediction, orchestration, optimization, simulation, and decision support, all applications that often operate quietly in the background. Their influence is measured less by what they produce than by how they improve the quality, consistency, and timing of institutional decisions.

Technology historians often note that the most transformative innovations eventually disappear into everyday life. Electricity became more consequential after it ceased to be remarkable. The internet reshaped nearly every industry only after it stopped being discussed as a feature. Mature technologies fade into the background because the systems built upon them become the focus.

The experience: LIAISON conference suggested that enrollment management may be approaching a similar point on the AI S-curve.

The early years of experimentation established that AI could perform individual tasks. The conversations in Baltimore were increasingly concerned with a different question: how intelligence can be distributed across an enrollment ecosystem to strengthen institutional capability. If that interpretation is correct, the next phase of AI adoption will likely be marked by fewer demonstrations of isolated features and more attention to the design of systems in which intelligence is simply expected to be present.