Artificial intelligence is no longer a novelty in higher education. Many institutions already use generative AI tools to help with teaching, research and daily admin. But the next phase of agentic AI adoption could change how universities actually operate.
According to Ashraf Davids, Solutions Manager for Digital Business Solutions at Datacentrix, the conversation is shifting from generative AI to agentic AI. The difference, he says, is not just technical. It is operational.
“Generative AI has shown value in improving productivity and supporting teaching,” says Davids. “But agentic AI is where institutions can see real change, especially in administration, compliance and decision-making.”
Generative AI tools mainly respond to prompts. They summarise text, draft emails or help with lecture notes. Their strength is speed and scale. Agentic AI, however, is designed to act. These systems can follow rules, make decisions and complete tasks across multiple platforms.
An agentic system could, for example, process an entire admissions workflow. It can check documents, flag missing information, notify applicants, update records and log each step for audit purposes. This moves AI from a support tool to an operational one.
Governance Becomes Critical as Agentic AI Adoption Grows
The shift brings clear benefits, but also new risks. Universities handle large volumes of sensitive data. As staff increasingly use public AI tools, concerns around data leakage and compliance are growing.
A recent global survey by the University of Melbourne and KPMG found that 72% of employees in emerging economies use AI regularly, compared to 49% in advanced economies. Yet fewer than half of organisations have formal AI governance policies.
Davids says this gap exposes institutions to legal and financial risk. Regulators are already taking action in some markets over data protection and copyright issues.
He argues that governance must be built into AI systems from the start. This includes role-based access, secure environments and clear audit trails.
As universities face rising student numbers and tighter budgets, agentic AI could help streamline higher education operations and improve support services. But the real shift, Davids notes, will depend on how well institutions balance automation with strong governance.
