A five-Step AI-Information problem solving (AI-IPS) model for critical, ethical, and responsible usage

The rapid advancement of generative artificial intelligence (AI) has fundamentally reshaped information problem-solving (IPS), creating a need for new frameworks in academic and professional contexts. This study proposes AI-IPS by analyzing scaffolded interaction logs, argumentative essays, and semi-structured interviews using inductive and deductive analysis. The participants were 124 undergraduate students from diverse academic backgrounds at a university. The proposed AI-IPS model comprises five key steps: defining the information problem, designing and refining prompts, analyzing and interpreting information, verifying and cross-checking evidence, and organizing and presenting findings. Our findings identify three essential competencies for effective AI-IPS: human-AI collaboration, independent thinking and critical reasoning, and information literacy. We suggest structured instructional scaffolding is required for ensuring the ethical and effective integration of advanced digital tools in education. This framework would equip students with critical reasoning skills and digital fluency, enabling them to navigate AI-generated content responsibly. Conceptually, AI-IPS is grounded in distributed cognition and socio-technical systems perspectives: cognition is accomplished across people, artifacts, and environments rather than residing solely in individuals, and effective performance requires the joint optimization of human and technical components.