<p>When generative artificial intelligence (GenAI) technologies emerged a few years ago, they were welcomed as innovative tools with the potential to transform teaching and learning. Yet now this initial optimism has been replaced by scepticism and concern over their role in perpetuating digital neocolonialism through linguistic hierarchies embedded in these tools. This study addresses these issues by examining how language teachers in the Global South perceive and negotiate pedagogical, ethical, and policy implications of generative artificial intelligence in English language education. Drawing on qualitative survey data and semi-structured interviews with 12 multilingual teachers across the Global South, the study explores their perceptions of GenAI’s pedagogical impact and their awareness of biases that AI tools can promote. Findings reveal teachers operating as policy arbiters in institutional vacuums, demonstrating variable but emerging critical awareness of how AI tools privilege certain English varieties, erase local linguistic practices, and reflect Global North epistemologies. The study argues that equitable AI integration requires systemic change: participatory policy frameworks, critical AI literacy, and fundamental redesign of tools to recognise linguistic and cultural diversity beyond dominant Western norms.</p>