Hi @NikSchet,
Thanks for reaching out!
The first version of Python to officially support M1 processors is 3.9.1, so I suggest you try setting up a new environment with that. Our dev team has recently started testing against 3.9, so PennyLane should run smoothly.
That said, if you are using any of the machine learning interfaces, you may run into trouble. In particular, tensorflow only supports up to Python 3.8, so I suggest to use the numpy/autograd interface, or pytorch.
Please let us know how it goes and if you run into further issues!