Short · Critical infrastructure · 2 min read

The vendor's remote access to your plant

Ask a plant manager who can connect into the control network from outside and you'll usually get a confident answer. Ask for the list and it takes a week. The list is longer than the answer.

How it accumulates

A vendor installs a system and needs support access. A jump box goes in. Then a cellular modem for the one site with bad connectivity. Then a second vendor with their own tool. Each one made sense on the day. Together they're a dozen unmanaged doors into the part of the business that physically does things.

What I'd want to know

  • Every path in, including the ones the vendors set up themselves.
  • Who holds credentials for each, and whether that's a person or a shared account at the vendor.
  • Whether access is always-on or enabled per session, and who enables it.
  • What's logged, and whether anyone reads it.
  • What the vendor's own security looks like, because their breach is your breach.

The controls that actually work in OT

Not the IT playbook copied over. Per-session access that an operator turns on, with a reason, and that turns itself off. Recording of what was done, not just that a login occurred. A single managed path rather than six, with the modems found and removed. And a conversation with each vendor about what happens if they're compromised, before it happens.

Why this is an executive issue

Remote access is how most serious OT incidents start. It's also usually a commercial relationship rather than a technical one, which means the fix needs someone who can renegotiate a support contract, not just someone who can configure a firewall.

If you run operations and aren't sure who can reach them from outside, I'd start there.

Explore: More on critical infrastructure ยท Work with me