Maybe. It is true that people who would have cheated in the past are now just using AI in addition to the previous means. But from my experience teaching, the number of students cheating is also increasing because of how prevalent AI has become and how easy it is to use it.

AI has made cheating more frictionless, which means that a student who might not have say used Chegg (requires some effort) or copied a friend (requires social interaction) in the past, can now just open a textbox and get a solution without much effort. LLMs have made cheating much easier, quicker, and safer (people regularly get caught using Chegg or copying other people, AI cheating can be much harder to detect). It is a huge temptation where the [short-term] benefits can greatly dwarf the risks.