Gradual Privacy Paradox in AI Fitness: Evidence of Privacy Satisficing from an Adult User Survey

AI-enabled fitness services collect continuous and sensitive data for monitoring and personalized feedback, which raises privacy and security concerns. Nevertheless, many users continue to engage with these services, suggesting a privacy–use tension. Using online survey data from 596 adults (age ≥ 18), this study examines AI fitness use from a privacy-satisficing perspective. We construct a Deviation index (standardized privacy concern minus standardized risk acceptance) and model high willingness to use AI fitness services with a parsimonious probability approach. Results indicate that continued use varies systematically across the Deviation spectrum. In logistic regression analyses, Deviation, perceived transparency and safety (Information Control Level, ICL), and privacy–convenience trade-off attitudes are associated with the likelihood of continued AI fitness use. Predicted probabilities vary gradually across the Deviation range. Overall, privacy concern and continued AI fitness use coexist in this sample, consistent with a bounded-rational privacy-satisficing interpretation.