Legacy Windows Machines on the Plant Floor: Managing the Ransomware Risk
Old Windows PCs run critical machines on many plant floors. Here is how manufacturers can isolate, harden and plan around them to cut ransomware risk.
Picture a quality lab at a mid-sized machine shop. The coordinate measuring machine (CMM) is calibrated, accurate and worth more than most of the cars in the parking lot. The PC running it still boots Windows 7, because the measurement software was never certified for anything newer. It's plugged into the same network as the front office, and nobody has logged into it as anything other than a local admin in years.
I see some version of this on almost every plant floor I walk. Legacy Windows in manufacturing isn't a sign that someone was lazy. It's what happens when a $250,000 machine is designed to run for twenty years and its control PC was designed to be supported for ten. The problem is that ransomware crews know this too, and an unpatched Windows box sitting on a flat network is one of the easiest ways into a plant.
You probably can't replace every old machine this quarter. What you can do is shrink the risk around them so that one infected laptop doesn't turn into a week of stopped production. Here's the approach my partner Mukesh and I use with manufacturers.
In this guide
- Why legacy Windows machines are still on your plant floor
- Why ransomware loves old Windows boxes
- Step 1: Find and rate every legacy machine
- Step 2: Wall them off from the rest of the network
- Step 3: Harden the machine itself
- Step 4: Control how people and vendors reach them
- Step 5: Plan for the day one of them fails
- Building a realistic replacement roadmap
- Frequently asked questions
Why legacy Windows machines are still on your plant floor
Most of the old Windows PCs I find in plants aren't general-purpose computers. They're part of a machine: the HMI on a press, the controller PC on a CNC, the software station for a CMM, a vision system or a label printer line. The equipment maker tested and certified a specific operating system, and changing it can void support or simply break the software.
Common reasons these machines never get upgraded:
- The software only runs on that OS. The vendor never released a newer version, or the vendor no longer exists.
- The upgrade is tied to a big capital purchase. A new controller can mean a retrofit that costs a meaningful slice of a new machine.
- Downtime is expensive. Nobody wants to be the person who took the line down to "just upgrade Windows."
- Nobody owns it. Operations thinks IT handles it, IT thinks it belongs to maintenance, and maintenance thinks the OEM handles it.
For reference, Windows XP stopped getting security updates in April 2014 and Windows 7 in January 2020. Mainstream Windows 10 reached end of support on October 14, 2025. Microsoft does sell paid Extended Security Updates (ESU) for Windows 10, and certain long-term servicing editions, such as Windows 10 IoT Enterprise LTSC 2021, are supported much longer (into 2032). If you're not sure which edition a machine runs, that's part of Step 1.
Why ransomware loves old Windows boxes
Legacy Windows security in manufacturing: the real weak points
An old Windows machine is risky for a few specific reasons, and knowing them tells you where to focus.
- Known, unpatched vulnerabilities. Once an OS is out of support, newly discovered flaws stay open forever. Attack tools for old flaws are widely available.
- Old network protocols. Many legacy machines still have SMBv1 turned on, the file-sharing protocol abused by WannaCry in 2017. That outbreak hit manufacturers hard because it spread on its own from machine to machine.
- Shared and weak credentials. Shop-floor PCs often run with a shared local admin account and a password that's been on a sticky note since installation.
- Limited security software. Many current antivirus and EDR products no longer support Windows 7 or XP, or support them with reduced features.
The bigger issue is position. If that old PC sits on the same flat network as email and web browsing, an attacker who lands on any office laptop can reach it. And because a stopped line costs real money every hour, attackers know manufacturers are under heavy pressure to pay.
Step 1: Find and rate every legacy machine
You can't protect what you haven't written down. Start with a simple spreadsheet, not a big software project.
What to capture for each machine
- Machine name, location and what equipment it runs
- Operating system and edition (check with winver on the machine)
- Whether it's on the network, and which network or VLAN
- Whether it has internet access
- Who supports it: your team, the OEM, or nobody
- What happens to production if it's down for a day
Rate the risk in three buckets
- High: On the main network, can reach the internet, or has remote access software installed.
- Medium: Networked but already on a separate segment with limited access.
- Low: Fully standalone, no network cable, files moved by a controlled process.
Walk the floor with your maintenance lead for this. In my experience, the spreadsheet and the reality rarely match on the first pass. You'll find a PC under a workbench that someone forgot about, or a cellular modem a vendor installed years ago.
Step 2: Wall them off from the rest of the network
Segmentation is the single biggest risk reducer for legacy machines. The goal is simple: an infected office PC should not be able to talk to a press HMI. I covered the broader approach in how to secure shop-floor OT networks without stopping production, so here I'll focus on what's specific to old Windows.
A practical segmentation checklist
- Put legacy machines on their own VLAN (or a few, grouped by function), behind a firewall, not just a switch.
- Default deny between zones. Allow only the specific connections the machine needs, such as one file server or one MES system on one port.
- No direct internet access. A legacy HMI almost never needs to browse the web. Block it at the firewall and log attempts.
- Block SMB and RDP from the office network into the legacy zone unless there's a documented reason.
- Use a transfer point for files. If engineers need to load programs, give them one controlled file share or jump host instead of open access.
Start in "monitor" mode if you're nervous. Log traffic for a couple of weeks, learn what the machines actually talk to, then write rules that match. That approach keeps operations on your side because nothing breaks on day one.
Step 3: Harden the machine itself
Once a machine is isolated, tighten it up. Do this during planned downtime, and take a full image backup first (see Step 5).
- Turn off what isn't needed. Disable SMBv1 if the application doesn't require it, along with unused services and network shares.
- Remove email clients and web browsers from production PCs where they aren't required. A machine that can't open email can't open a phishing attachment.
- Change shared passwords. Give each machine a unique local admin password and store it in a password manager, not on the monitor.
- Use standard user accounts for operators. The operator shouldn't need admin rights to run a job.
- Control USB drives. USB sticks are one of the most common ways malware reaches air-gapped machines. Use dedicated, scanned drives or a scanning kiosk.
- Consider application allowlisting. On a machine that runs one or two programs, allowlisting can block anything else from running. Test carefully with the OEM software.
- Check security software support. Ask your EDR vendor which legacy OS versions are still supported and with what features. Don't assume the agent you use on laptops works on a Windows 7 HMI.
Apply whatever patches the machine is still eligible for. Many old controller PCs are missing updates that were released before end of support, simply because nobody installed them.
Step 4: Control how people and vendors reach them
Remote access is where I see the most avoidable risk. OEMs often install their own remote tools to support a machine, and those connections tend to outlive the project they were created for.
Tighten remote access
- Find every remote access path. Look for remote desktop tools, VPN clients, cellular modems and anything listening for inbound connections.
- Route vendor access through one controlled gateway with multi-factor authentication, individual accounts and logging.
- Turn vendor access on when needed, off when done. Scheduled or approved sessions beat always-on tunnels.
- Never expose RDP directly to the internet. If a legacy machine has a public RDP port, treat that as an emergency.
If you're a defense supplier, how you handle this also matters for compliance. Machines that touch controlled unclassified information are in scope, and our post on CMMC Level 1 vs. Level 2 explains how scope works. Your assessor will want to see how unsupported systems are isolated and documented, so confirm the details with them.
Step 5: Plan for the day one of them fails
Even with strong controls, assume one of these machines will eventually be hit by malware, a failed hard drive or a power surge. Old hardware fails, and replacement parts get harder to find every year.
A recovery kit for each critical legacy machine
- A full disk image, tested by restoring it to spare hardware at least once
- Offline copies of installers, license keys, drivers and machine configuration files
- The OEM support contact and any support contract number
- Spare hardware for the most critical machines, or a known source for compatible parts
- A written restore procedure that someone other than the original technician can follow
Store at least one copy of all of this offline or in immutable storage. If ransomware encrypts your file server, a backup sitting on that same server won't help you.
Building a realistic replacement roadmap
Isolation and hardening buy you time. They don't make an unsupported OS safe forever, so pair them with a plan to retire the riskiest machines.
- Rank machines by risk and business impact using your Step 1 spreadsheet.
- Ask each OEM directly whether a supported OS or controller upgrade exists, and get a written quote.
- Look at your options for each machine: upgrade the software, replace the controller PC, use Windows 10 ESU as a bridge where eligible, move to an IoT or LTSC edition the OEM supports, or replace the equipment.
- Tie upgrades to planned capital spending and scheduled shutdowns, like the holiday week or annual maintenance.
- Review the list every year. Lifecycle dates change, and Microsoft's lifecycle pages are the source of truth.
On cost: Windows 10 ESU is priced per device, and commercial pricing doubles each year it's renewed, so it's usually a bridge rather than a long-term plan. Check Microsoft's current pricing and eligibility before you budget.
This roadmap is also a great conversation starter with your CFO or owner. A one-page list that says "these five machines are our biggest exposure, and here's what fixing them costs" gets better results than a general request for security budget.
Frequently asked questions
Is it safe to run Windows 7 on a machine that isn't connected to the internet?
It's much safer, but not automatically safe. Malware can still arrive by USB drive, from a vendor laptop or from another machine on the same network. Isolate it, control USB use, and keep a tested image backup.
Can I still get security updates for Windows 7 or XP?
Not for general business use. Windows 7's paid extended updates ended in early 2023, and XP support ended in 2014. Some embedded editions had different timelines, so check the exact edition and Microsoft's lifecycle page.
Do I have to replace machines running Windows 10 now that support ended?
Not necessarily right away. Eligible businesses can buy Windows 10 Extended Security Updates for a limited number of years, and some IoT Enterprise LTSC editions are supported until 2032. The key is knowing which edition each machine runs and having a plan before coverage runs out.
Will CMMC or my cyber insurer accept legacy Windows machines?
Often they can be accepted if they're documented, isolated and protected with compensating controls, but it depends on the assessor, the insurer and the specific requirement. Expect questions about how the machines are segmented and who can reach them. Confirm with your assessor, insurer or attorney before you rely on any approach.
How do I protect a legacy HMI without breaking production?
Most of the protection happens around the machine rather than on it. Segment it behind a firewall, block internet access, restrict remote access and USB use, and make changes on the machine itself only during planned downtime with a backup in hand.
Get a second set of eyes on your plant floor
If you're not sure how many legacy machines you have, or how exposed they are, that's a good place to start. Our free cybersecurity assessment looks at your network layout, remote access and the systems that would hurt the most if they went down. You'll come away with a short, prioritized list you can act on, whether or not you work with us afterward.
Related reading
- How to Secure Shop-Floor (OT) Networks Without Stopping Production
- CMMC Level 1 vs. Level 2: Which One Does Your Manufacturing Shop Need?
- Cybersecurity for Roanoke Small Businesses: 10 Steps to Protect Your Company
Harsha is a co-founder of SoftwarePac Technologies in Roanoke, VA, and a network engineer with 15 years of experience.




