Illinois Health Officials Use ChatGPT to Investigate Salmonella Outbreak Linked to County Fair
Key Takeaways
- ▸Illinois health officials used ChatGPT to investigate a Salmonella outbreak affecting 13 people linked to a county fair's beer tent
- ▸The outbreak was traced to beer cans stored in an unsanitary makeshift cooler made from non-food-grade drainage tile that was never properly cleaned
- ▸Investigators asked ChatGPT questions about bacterial growth and similar outbreak precedents after the contaminated cooler had been disposed of
Summary
Health officials in Brown County, Illinois, turned to ChatGPT to help investigate a puzzling Salmonella outbreak connected to a county fair in August 2024. The outbreak affected 13 people across five counties, all of whom had attended the Brown County fair and consumed canned beer from the event's sole beer tent. Investigators discovered the beer was stored in a makeshift cooler—a 10-foot length of non-food-grade corrugated black plastic farm drainage tile that was never properly cleaned or drained throughout the fair's duration.
With the contaminated cooler disposed of after the fair, health investigators faced challenges confirming their hypothesis that Salmonella had contaminated the beer cans through the unsanitary ice and handling practices. According to a report in the CDC's Morbidity and Mortality Weekly Report, lead investigator Katherine Houser and her team consulted ChatGPT, asking questions such as whether S. Agbeni could grow in an improperly drained cooler and whether similar outbreaks had been documented in scientific literature. While some of these questions could have been answered through traditional database searches like PubMed, the officials sought AI assistance for validation.
The case highlights both the potential and limitations of using AI chatbots in public health investigations. While ChatGPT provided reassurance to the investigators about their hypothesis, the report notes that whether the AI was actually helpful remains unclear. The incident raises questions about the appropriate role of generative AI tools in epidemiological work, particularly when traditional research methods and expert consultation remain available and may be more reliable for critical public health decisions.
- The effectiveness of using AI chatbots for epidemiological investigations remains questionable, as traditional research databases could have provided similar information
