There is a dearth of evidence on how adopting AI technologies relates to L2 reading comprehension. To fill the gap, our study adopted a quasi-experimental pretest-posttest design to explore the impact of AI-enhanced reading instruction on English students’ reading comprehension. Seventy-two English as foreign language (EFL) learners were split into an experimental and a control group. Both groups completed a Test of English as a Foreign Language (TOEFL)-style reading test as a pre and posttest. Over eight weeks, the experimental group adopted AI tools (ChatGPT, reading bot, and Copilot) to complete reading tasks, while the control group received traditional instruction with TOEFL preparation coursebooks. The results of an independent-samples t-test and ANCOVA revealed that students in the experimental group obtained significantly higher scores in reading comprehension than the control group. Because the intervention integrated AI tools with structured reading strategies, including summarization, verification, vocabulary exploration, and questioning, the observed improvements should be interpreted as reflecting the combined instructional approach rather than the independent effect of AI technologies alone. Our results also proposed that students with higher AI familiarity were disposed to demonstrate larger gains, but this difference across familiarity sub-groups was not statistically significant. Our study offers practical implications for teachers to use AI tools more in reading instruction.