This systematic review asks how AI literacy among language teachers in higher education is defined, how it's developed through professional learning, how it's assessed, and what outcomes and gaps studies report. It answers through a thematic synthesis of 32 empirical and conceptual studies published between December 2022 and March 2026, read through critical policy and sociomaterial lenses.
- Most studies define AI literacy with competency-based, multi-dimensional models, while its critical and domain-specific dimensions remain underdeveloped.
- Professional development for these teachers is often unstructured and poorly planned.
- Assessment leans heavily on self-report tools and gives little attention to ethical and critical skills.
- Teachers report individual gains such as greater confidence and new teaching practices, but limited institutional support and unequal access to resources can weaken those gains.
- Unclear ethical guidelines and shared or blurred responsibility for AI literacy also work against positive outcomes.
The authors conclude that AI literacy among these teachers is built mostly through personal effort and ad-hoc experimentation. The institutional governance structures that could make it durable and equitable are still underdeveloped. In their reading, the evidence shows that sustainable AI literacy depends on structural governance from institutions, and that individual upskilling can't carry it alone. They call for longitudinal and linguistically diverse research that pays closer attention to governance.
