2026 ASEE Annual Conference & Exposition

This complete research paper investigates the early patterns of artificial intelligence (AI) engagement among first-year engineering students prior to formal instruction or policy intervention. A survey administered near the end of the Fall 2025 semester collected self-reported data on frequency and purpose of AI use, trust in AI-generated outputs, verification behaviors, perceived benefits and risks, and engineering identity. A total of 818 students consented to participate. Results show that AI use is nearly ubiquitous among first-year engineering students and is frequently embedded in core academic tasks, including studying concepts, brainstorming, and problem solving. Participants report moderate trust in AI accuracy, lower confidence in its ethicality and potential for bias, and varied verification practices, while a smaller subgroup rejects AI as a learning tool and exhibits significantly lower trust, suggesting principled resistance rather than lack of exposure. At the same time, all participants express strong concern about over-reliance on AI, and high agreement that AI should be used responsibly by professional engineers. These findings suggest that first-year students are already navigating complex relationships with AI, balancing perceived learning benefits with concerns about independence and professional responsibility. The study establishes an important baseline for AI use in engineering education and highlights the need for instructional approaches that emphasize verification, ethical judgment, and reflective AI use as components of professional formation. This work provides a foundation for longitudinal research examining how structured AI literacy instruction influences student practices, trust, and identity over time.