Governing generative AI in higher education: a global Delphi study on policy and practice

Synthesis: Crompton and a large international panel used a Delphi technique and collective writing to gather expert perspectives from 22 countries/locations across six continents, producing a consensus-driven higher education GenAI policy framework with eight core areas: academic integrity, ethical and responsible use, privacy and protection, equitable access, GenAI literacy, integration strategy, human oversight and accountability, and institutional support and infrastructure. A complementary six-part mechanism — a dedicated GenAI committee, scheduled policy reviews, ongoing professional development, stakeholder communication, impact evaluation, and monitoring of external developments — is proposed to keep policies current. Grounded in the Socio-Ecological Technology Integration (SETI) framework, the study positions policies as enabling structures within an interconnected institutional ecosystem rather than isolated rules.

Core Finding

Effective GenAI governance in higher education requires a coordinated, consensus-driven framework that integrates policy, pedagogy, ethics, and infrastructure, with academic integrity as a central regulatory anchor whose practical implementation depends on GenAI literacy and ethical reasoning. Policies alone accomplish little without institutional support, so universities should move beyond reactive, localized restrictions toward proactive, adaptable governance that treats GenAI literacy as a cross-cutting capability.

The Eight-Part Governance Framework

The panel converged on eight core areas that collectively function as an interconnected institutional ecosystem. Academic integrity operates as a central regulatory anchor, with strong consensus on pairing a strict policy backbone with educational guidelines. Ethical and responsible use (referenced by two-thirds of panelists) extends beyond integrity into honesty, transparency, fairness, and broader societal impacts, while privacy and data protection emerged as critical themes for over half the panel, aligned with standards such as GDPR and FERPA. Equitable access (one-third of panelists) addresses digital divides and algorithmic bias, and GenAI literacy (over half of responses) is framed as a foundational graduate capability that empowers critical and responsible use.

Integration, Human Oversight, and Institutional Support

Panelists strongly advocated proactive integration of GenAI into teaching and learning — "support, not replace" — favoring process-focused and oral Assessments that assess higher-order skills rather than outputs a model could generate. Roughly 40% of panelists discussed pedagogical strategies or assessment in the context of GenAI. Human oversight and accountability must remain central, with half the panel insisting that significant GenAI outputs be reviewed and validated by a human and that accountability never be abdicated to an algorithm. Institutional support and infrastructure serve as the foundational enabler, with panelists recommending dedicated GenAI committees or support offices, senior leadership endorsement, and iterative investment in infrastructure.

Keeping Policies Current

A six-part review mechanism ensures policies remain current as GenAI evolves: a dedicated multidisciplinary GenAI governance committee (recommended by over 50% of panelists), scheduled policy review cycles (half the panel), ongoing professional development, communication with all Stakeholders, evaluation of effectiveness and impact, and monitoring of external developments. These processes treat policy maintenance as an ongoing institutional mechanism with clear ownership rather than a one-time task.

Relevance to the wiki

This paper provides the wiki's most authoritative, globally consensus-driven account of Governance and Educational Policy AI for Generative AI in Higher Ed, offering a concrete blueprint that connects Academic Integrity, Ethics, Privacy, and AI Literacy. Its insistence on human oversight directly informs the wiki's treatment of human Agency and accountability, while the equity dimension speaks to the wiki's concern with fair and inclusive adoption. For practitioners, the eight-area framework and six-part review mechanism are actionable templates for institutional policy design.

Connected Concepts

Connected Articles

Citation

Crompton, H., Burke, D., Nickel, C., Bozkurt, A., Miao, F., Sharples, M., et al. (2026). Governing generative AI in higher education: a global Delphi study on policy and practice. International Journal of Educational Technology in Higher Education.

View Source View Changes Edit