EC2 instances launched via Terraform do not pass the status checks or have insufficient data

Question

I use Terraform to launch 6 new m3.medium EC2 instances via different Auto Scaling Groups.

4 out of 6 have state: running but status_check: Insufficient data or 1/2 checks passed. Basically only 2 launch correctly.

It seems really strange behaviour and I was wondering if anyone else encountered it.

Does it have to do with some limit of this type in specific region/AZ ?

Any pointer would be appreciated.