Azure Application Gateway has 2 different SKUs, V1 and V2. the V2 SKU has optional autoscaling that can be configured to help the Application Gateway handle a dynamic load. You can choose to have this disabled, or enabled with defined limits on the scaling (Automatic Scaling Yes/No). If you have it enabled, you define the minimum instances (Minimum scale unit) and maximum instances (Maximum scale unit).
An instance of Application Gateway is a bit complicated, and is fully explained here. Billing for the V2 SKU is complicated, and that doc outlines how it happens.
You should change these default settings if you want autoscaling on your application gateway to handle the load. They can be disabled or enabled at any time, so this is not something that you need to decide on now.
Typically, people use autoscaling when they have intermittent high traffic loads that the normal instance count cannot handle, and they want to save money during periods of low traffic. This is something you can monitor on your Application Gateway and set up at a later time when you have a better understanding of the traffic loads of your applications.
I am unsure if this is directly covered in the az900 exam. The V2 SKU is fairly new, so it might not be on the test. Please let me know if you have any additional questions.