How do you roll out Generative AI across 30+ countries?
A licence everyone has → a tool the majority actually use
The problem
CHAI bought access to generative AI tools for teams in 30+ countries. That decision took an afternoon. Getting people to use them was the real problem: thousands of staff, each in a different healthcare system, each starting from a different level of technical comfort, each choosing every day whether to open the tool or fall back on the way they already knew. Some workflows touched sensitive health data, which raised the stakes for anyone trying it for the first time. The brief read like a software rollout. The work was behaviour change at organisational scale.
What I did
Ran it in stages instead of one launch. Progressive pilots to find which use cases actually stuck. Internal resources and training so people had a low-stakes way in. Governance for the things that matter, sensitive data especially, so a first-timer could use AI without worrying they'd break something. That took the organisation from a handful of early adopters to the majority of staff. The question has since shifted from 'is anyone using this' to 'what are the best users doing,' surfacing those use cases and scaling them across teams. Part-time engagement with CHAI's US HQ, ongoing.
Outcomes
What I learned
AI licences are like gym memberships. Leadership buys them in an afternoon. Using them is a behaviour thousands of people have to choose every single day.
Most of this job was human change management wearing an AI systems-design hat.
Governance isn't what slows adoption down. Done right, it's what lets a nervous first-timer try AI without fear of breaking something.
Got a similar problem?
Tell me what you're working on. I'll tell you if it's a fit.
Prefer email? workwithakshatkharbanda@gmail.com →