AI sprawl is one of those problems that's easy to miss because every individual tool adoption seems reasonable. The real challenge starts when different teams are sending sensitive data to different models with no shared governance or visibility.
A practical approach we've taken at IT Path Solutions is to standardize on a small set of approved AI platforms while keeping model access flexible behind the scenes. That gives teams room to experiment without creating dozens of disconnected AI workflows that become difficult to secure, audit, and maintain later.
This is a great example of exactly the balance I was trying to get at,
governance without killing experimentation. Standardizing on a small
set of approved platforms while keeping model access flexible sounds
like it solves the visibility problem without making teams feel locked
out of trying new things.
Curious how you handle the review process when a team wants to add a
new model to that approved set, is it a lightweight request, or more
formal?
Mateo Ruiz
Senior Tech Consultant
AI sprawl is one of those problems that's easy to miss because every individual tool adoption seems reasonable. The real challenge starts when different teams are sending sensitive data to different models with no shared governance or visibility.
A practical approach we've taken at IT Path Solutions is to standardize on a small set of approved AI platforms while keeping model access flexible behind the scenes. That gives teams room to experiment without creating dozens of disconnected AI workflows that become difficult to secure, audit, and maintain later.