In 2023, psychostimulants, mostly methamphetamine, were involved in 35,982 overdose deaths in the United States. Methamphetamine is involved in approximately one in three fatal overdoses, underscoring its role in the ongoing overdose crisis. Until 2022, when the methamphetamine overdose ICD-10-CM code T43.65 was published, capturing nonfatal methamphetamine overdose trends was limited by the lack of a specific methamphetamine poisoning or overdose diagnostic code. In 2024, the Centers for Disease Control and Prevention (CDC) Drug Overdose Surveillance and Epidemiology (DOSE) program released a nonfatal methamphetamine overdose syndrome definition that incorporates both discharge diagnosis codes and analyses of chief complaint text from electronic health record data. This syndrome definition can now be used to gain a more detailed understanding of nonfatal methamphetamine overdose patterns and trends over time to guide public health response. This includes identifying which signs and symptoms are most frequently reported by patients receiving emergency medical treatment for methamphetamine-involved overdoses. Integrating generative artificial intelligence (GenAI) tools such as ChatGPT models into public health surveillance can streamline natural language processing tasks, improving both the timeliness and accuracy of detecting public health trends. Applied to nonfatal overdose surveillance, these tools have the potential to accelerate the development and validation of syndrome definitions, allowing for more rapid identification of emerging drug issues and refinement of existing definitions used for ongoing monitoring. This dissertation includes three interrelated studies. The first proposes a framework to apply GenAI-based natural language processing methods to enhance validation of the DOSE suspected methamphetamine-involved overdose syndrome definition. The second describes demographic characteristics and temporal trends among suspected nonfatal methamphetamine overdose emergency department visits reported to the CDC National Syndromic Surveillance Program (NSSP), as well as the frequency of key syndrome query terms appearing in chief complaint text. The final paper presents a commentary that explores the appropriate and ethical integration of GenAI tools into public health surveillance systems and offers recommendations for their future use. Among ED visits reported to NSSP, 22,192 suspected nonfatal methamphetamine-involved overdoses were identified from October 2022 through May 2025. Among these emergency department visits, patients were most frequently male (70.4%), non-Hispanic White (66.7%), aged 25–44 years (61.2%), and treated in facilities located in the southern United States (48.3%). Nonfatal methamphetamine overdose rates were highest in June 2023, 1.22 ED visits per 10,000 and lowest in January 2025, 0.63 ED visits per 10,000. Collectively, this dissertation introduces an innovative framework for integrating GenAI into the DOSE syndrome validation process through the CDC ChatBot, a secure application that merges ChatGPT models with appropriate safeguards for sensitive data analysis. This framework can be refined to support the validation and revision of other DOSE syndrome definitions, as well as the development of future syndromic surveillance tools. The final study, a commentary, extends these findings by discussing critical considerations for the appropriate and ethical use of GenAI in public health surveillance and offering recommendations to guide its implementation across the field. Together, these studies demonstrate how GenAI can strengthen public health surveillance systems, improve the identification of nonfatal methamphetamine-involved overdoses, and inform data-driven interventions for the most affected populations.