Did ChatGPT get worse somehow?

I know there's a cancel wave, and I personally have also canceled my subscription, but I've still been using it due to my location and VPN usage making some other models a bit of a hassle to access. However, prior to and after my cancellation, I noticed that the model keeps trying to force answers based on its existing knowledge, instead of verifying information. Is this something new, or am I tripping? It feels infinitely worse than it used to. This is with my tone set to Professional and having this custom instruction: Don't be condescending. Get to the point.