Extending the affective process and strategic engagement framework to address writer’s block in Ethiopian English as a foreign language contexts through critical artificial intelligence literacy and ethical generative artificial intelligence in low-r

Abstract Writer’s block remains a serious obstacle for undergraduate English as foreign language writers, particularly in low-resource Global South settings, where writing anxiety, linguistic insecurity, and unreliable internet access intensify the challenge. Researchers have shown growing interest in generative artificial intelligence tools such as ChatGPT as writing support resources, yet existing scholarship has not yet explained how EFL learners in such settings engage with these tools critically and purposefully. This study asked three questions. How do undergraduate EFL writers at a public university in Ethiopia use generative AI to manage writer’s block? What does that engagement reveal about their development of critical AI literacy? The researcher conducted an interpretive multiple case study with ten undergraduate students at Bahir Dar University across one academic semester. Data sources included semi-structured interviews, weekly reflective journals, anonymized ChatGPT conversation logs, successive writing drafts, and classroom observations. The researcher analyzed the data thematically and strengthened trustworthiness through member checking. The analysis revealed a clear developmental pattern. Most participants began by relying on generative AI without critical evaluation, and then gradually shifted toward more selective and purposeful use. They identified the cultural and linguistic limitations of AI-generated text, reasserted their authorial voice against perceived Western biases in the output, and applied deliberate prompting strategies while rejecting content that did not reflect their intended meaning. Participants also reported reduced writing anxiety through intentional and ethically considered use of generative AI. Unreliable internet access, however, remained a persistent structural barrier that limited the depth of their engagement. These findings extend the APSE framework for critical AI literacy through two additional layers. The affective layer addresses anxiety reduction, confidence restoration, and identity resilience. The equity layer addresses connectivity constraints, linguistic hierarchies, and resistance to Anglocentric norms. The study recommends context-sensitive teaching approaches that balance generative AI support with student agency, cultural relevance, and equitable access to technology.