I will be curious to see if it has any impact positive or negative over a couple of years.

Will the code be more secure since the AI does not make the mistakes humans do?

Or will the code, not well enough understood by the employees, exposes exploits that would not be there?

Will it change average up time?