This study explores university students’ perceptions of using AI-based applications
specifically ChatGPT, Grammarly, and QuillBot to develop their English writing
skills. Responding to the lack of research on students' subjective experiences with
AI tools in the Indonesian higher education context, this study adopted a qualitative
case study design. Data were collected through open-ended questionnaires and
semi-structured interviews with five English Education students who frequently use
AI writing applications. Thematic analysis following Miles, Huberman, and
Saldana’s (2014) framework revealed four key findings: (1) students frequently
used AI tools in all stages of writing, particularly for grammar correction,
paraphrasing, and idea generation; (2) students perceived these tools as helpful for
improving structure, clarity, speed, and confidence; (3) challenges included
overdependence, inaccurate output, and reduced critical thinking; and (4) students
preferred a blended approach using AI tools alongside traditional teacher feedback
for more effective and meaningful learning. These results confirm the relevance of
the Technology Acceptance Model (TAM), Writing Process Theory, and CALL
framework in understanding how AI applications influence writing development in
EFL contexts. The study suggests that while AI tools offer significant support, their
optimal impact depends on critical, guided use integrated with human instruction.

Keywords: AI-based mobile applications, student perception, academic writing,
EFL, Grammarly, ChatGPT, QuillBot