TRACKING GENERATIVE AI ADOPTION IN A CULTURALLY DYNAMIC SETTING: DAILY GOOGLE TRENDS EVIDENCE FROM THE UAE

Understanding the temporal dynamics and cultural determinants of generative AI adoption is essential for informed policymaking and effective deployment. This study presents a high-frequency, culture-aware analysis of public interest in ChatGPT within the United Arab Emirates (UAE), utilizing daily Google Trends data from April 2024 to May 2025. The objective is to dissect temporal patterns, assess volatility and event responsiveness, and benchmark predictive models for short-term forecasting. The methodology integrates time series decomposition (STL), volatility modeling (GARCH), anomaly detection, and autocorrelation analysis to uncover behavioral drivers. Special attention is given to cultural phenomena—particularly the effect of Ramadan—on usage patterns. A comparative forecasting experiment evaluates four models: SARIMA, Prophet, N-HiTS, and a feature-engineered Random Forest. The analysis reveals: (1) a sustained upward trend in ChatGPT interest; (2) a strong weekly cycle centered on the workweek; (3) event-driven volatility spikes linked to model releases and holidays; (4) a culturally induced drop in engagement during Ramadan; and (5) Random Forest outperforming all other models with a MAPE of 10.65%. This is the first empirical study to combine high-frequency trend analysis, cultural calendar effects, volatility modeling, and machine learning forecasting in the context of AI adoption in the MENA region. It introduces a novel, interpretable framework for real-time monitoring and forecasting of digital innovation diffusion, offering strategic insights for policymakers, educators, and AI developers operating in culturally dynamic settings.