CER: Helping Adults Make Safe and Informed Decisions About Artificial Intelligence Across the Lifespan

Artificial intelligence (AI) is becoming part of everyday life. People use it to search for information, complete forms, manage finances, communicate with others, and make decisions. However, AI-based systems have the capacity to produce inaccurate, misleading, or fabricated information while presenting the information in a way that appears confident and convincing. Many adults must decide whether to trust, question, verify, or reject these outputs without understanding how this technology works or recognizing its limitations. Poor decisions can increase vulnerability to hallucinated and incorrect outputs, financial loss, fraud; however, excessive skepticism may prevent people from benefiting from useful technologies. This project seeks to understand how adults of different ages learn to make safe and informed decisions when using AI-based systems. The knowledge gained will help educators and community organizations develop effective learning opportunities that strengthen public understanding of AI, promote responsible use of emerging technologies, and improve the ability of adults to navigate an increasingly digital society. This project investigates how adults develop trust calibration, judgment, and safe use reasoning when interacting with generative AI-based systems under conditions of uncertainty. Using a mixed-methods design-based research approach, the team of researchers will conduct iterative cycles of instructional design, implementation, analysis, and refinement over three years. Emerging, middle-aged, and older adults will participate in community-based workshops and short courses offered through a community education partner. Participants will engage with realistic scenarios involving information seeking and decision making and will determine when to rely on, verify, defer, or reject AI-generated outputs. The study will examine how instructional features such as uncertainty cues, verification supports, and reflective prompts can influence reasoning processes and reliance decisions. Quantitative and qualitative evidence will be integrated to identify learning mechanisms and develop transferable design principles for adult AI literacy. The project will advance understanding of how adults learn to reason about AI and will provide evidence-based guidance for designing educational experiences that support appropriate trust and safer use of AI technologies across adulthood. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.