Thanks for your response.

The current path doesn't form a close loop but it has an incoming and outgoing terminal with same amount of current. What's more peculiar is that when I remove all the magnetic boundaries, it simulates the structure without any error (meaning that the current path on its own is fine), its only when I have the magnetic boundary conditions that I get the 'current leak into the air' problem. I think its some interaction between the tangential boundary conditions and the current boundary conditions during simulation. It gives all checks post validations, its only during the simulation that the current leakage error is reported.

I don't  know how to fix this