Secondary Students’ Perceptions of AI-Assisted Learning: A Mixed-Methods Study of Independence, Motivation, Critical Evaluation, and Socratic-Inspired Filters

Generative artificial intelligence is increasingly used by secondary students to obtain explanations, explore ideas, and complete academic tasks. Its educational value, however, depends on whether it supports reasoning or replaces independent effort. This study examined students’ perceptions of learning independence, motivation, critical evaluation, and their use of the Socratic-inspired filters of truth, goodness, and usefulness. A quantitative-dominant, cross-sectional convergent mixed-methods survey involved 26 students from Grades 7–12 who participated voluntarily at an Indonesian secondary school. Data were collected from December 2025 to January 2026 through 13 five-point Likert-scale items and three open-ended questions. Quantitative responses were analyzed descriptively, while written responses were examined thematically and integrated during interpretation. AI was perceived as more supportive of topic exploration than of general learning motivation. In separate items, 65.4% reported greater exploration and 65.4% greater questioning, while 73.1% disagreed that they accepted AI responses without evaluating accuracy. Positive responses reached 69.2% for truth, 61.5% for usefulness, and 46.2% for goodness. Open-ended responses highlighted clarification and easier access, alongside concerns about reduced effort and dependence. Overall, participants described AI as a useful learning support when accompanied by independent attempts, verification, ethical consideration, and explanation.