Navigating Challenges When Adopting New Technologies

Building trust in AI tools requires transparency from the ground up. A tool designed to predict lifetime cardiovascular risk, for example, needs clear explanations of how it was built, how its underlying data is validated, and why its outputs can be trusted by both consumers and physicians. Broad adoption is unlikely to happen quickly — it will depend on published evidence, transparency about methodology, and potentially FDA approval or endorsement from recognized industry groups.

At the team level, the path forward runs through experimentation. Encouraging people to explore what tools like Codex can do opens up possibilities, but moving from experiment to production requires rigor around data security, privacy, coding standards, and validation. As AI takes on more analytical work, roles in analytics may shift toward data structure and validation rather than disappear — and leaders have a responsibility to explain that evolution clearly, so team members understand how their work is changing rather than fearing they’re being replaced.