If the three new employees can communicate with each other's devices, that strongly suggests their local Layer 2/Layer 3 connectivity is working (same VLAN/subnet, switching is fine, and they have IP addressing that allows local communication). The fact that they cannot access company resources (often on other subnets, servers, or the internet via internal routing) points to a problem with how these new endpoints are being addressed or routed beyond their local segment.
The most likely shared root cause affecting only "new employees" is DHCP scope/pool configuration. If the DHCP pool is exhausted, mis-scoped, or handing out incorrect options (such as an incorrect default gateway, wrong subnet mask, or an address range tied to an isolated VLAN), users could still talk locally but fail to reach resources outside their subnet. Modifying the DHCP server pool (expanding the scope, correcting the scope network, or ensuring the correct options are assigned) would resolve this with minimal changes.
A gateway routing table issue would typically affect many users, not just three new ones. DNS misconfiguration on the router would more commonly cause name-resolution problems, not block all resource access (and wouldn't be isolated to only new employees). A workstation firewall would not usually prevent access to "company resources" in a way that consistently affects three new devices while still allowing peer connectivity.