In theory if there are no security holes, a user account can only mess up its own account.
Note that what steps you want to take will really depend on who these users are and what you want to achieve. There’s a vast chasm between allowing in, say, friends or colleagues, vs. letting random people on the internet access it. The latter will mean someone will intentionally look for exploits, which means e.g. regularly applying security updates becomes far more pressing.
If you are letting in random strangers, I’d look into only giving them access within a separate container or ideally virtual machine per user as an extra precaution unless what you’re making available is very stripped down.
If it’s not connected to the internet, that certainly does take away a significant cause of problems. Good luck