We banned AI tools for a quarter. Usage moved to phones.
○ Didn't work
What happened
After a scare about customer data in prompts, leadership blocked every AI website on company machines. For three months the official numbers looked perfect. Then an expense audit found a dozen personal subscriptions, and a quick anonymous poll showed most of the team was using AI on their own phones, with no rules at all.
What it cost
The ban itself cost nothing. Undoing it took a quarter of trust.
What we fixed
We shipped one approved tool with a business data agreement, a one-sentence rule for customer data, and a no-blame amnesty to list what people actually used.
The lessonA ban without an approved alternative just moves the risk out of sight.
Reality numbers
- Team
- 51–200 people
- Monthly AI spend
- don't know
- Still running after 90 days
- rolled back