FORSBERG, L. Instruction Alignment and Risk Calibration in Large Language Models for Safe Human-AI Interaction. Journal of Computer Technology and Software, v. 4, n. 6, 30 June2025.