Generative AI Usage and Trust Among Business Users

This research report examines the use of generative AI for analytical work among over 100 business users, focusing on their trust levels and verification practices. Despite acknowledging the unreliability of generative AI outputs, users continue to utilize these tools for high-stakes decisions and reports. The findings indicate that 74% of users have made decisions based on incorrect AI-generated numbers, with a significant portion doing so recently. The report highlights a notable gap between usage and trust, with users often verifying AI outputs through methods that may not effectively catch errors. Additionally, a majority of users believe that generative AI tools like ChatGPT could replace traditional business intelligence tools, reflecting a shift in analytical practices. The report concludes that while users are aware of the risks associated with generative AI, they prefer its speed and familiarity over traditional methods, despite the potential for inaccuracies in their work.