How To Extend Your Firewall Protection To Remote Workers
November 11, 2026 ·
The perimeter is where the people are
Pre-COVID, remote access was the occasional requirement. Today, that is the norm, not the exception. As of last year, almost 50% of UK workers were either remote or hybrid, and in 2026 that figure has likely gone over the 50% mark. The VPN scramble of 2020 was meant to be a temporary fix. It's now permanent, it's here to stay.
So the perimeter is no longer at the firewall. The perimeter is where the people are and where the devices are, whether that's at home, a client site, on the train or in a coffee shop.
What's the problem with VPNs?
In a nutshell, traditional VPNs were built to extend the network, not to secure devices. Once you're in, you're in. Stats from Q3 2025 showed that for 48% of ransomware attacks, the initial access came through a VPN, and that is growing.
Two things we repeatedly see in breaches today are stolen credentials and exploitation of remote access infrastructure, and over the last year we've seen serious vulnerabilities in remote access products from vendors across the market. The Akira ransomware group is a good example: they target internet-facing VPNs, use stolen credentials for the initial foothold, then go after servers, virtual infrastructure and backups, all in a very short window.
AI is making this even easier. In one demonstration, an attacker found an exposed SonicWall VPN, picked a vulnerability from CISA's known exploited vulnerabilities list and asked their AI coding agent to build an exploit. In a few minutes it was stealing another connected user's session cookie, giving free access to the local network. If you have VPN services in place today and you watch the logs, you'll see there's pretty much a persistent attack going on against them.
There's a compliance angle too. Once you get to Cyber Essentials Plus, any device touching organisation data is in scope and must be protected, so you can't rely on the home router. ISO 27001 goes further, naming MFA, least privilege access and device controls explicitly. A VPN covers the encrypted tunnel, but it doesn't cover access control and it doesn't cover logging.
AI on both sides
Users are increasingly adopting SaaS and AI tools directly, and uploading company data into an unmanaged AI service can create data leakage and compliance risks. AI itself isn't the problem, it's unmanaged AI. If you don't know what your users are accessing, how can you secure it?
Attackers are using AI too, less to invent new types of attack and more to make existing ones faster, cheaper and easier to scale, from tailored phishing to deepfakes of your CEO. Agentic AI hacking will collapse the attack timeline from weeks to minutes. The same technology helps defenders spot anomalies and shadow IT. AI doesn't replace the security team. It helps the team see more, understand faster and respond sooner.
Practical steps everyone should consider
- Keep the basics strong. End-of-life hardware and software may remain permanently unpatched.
- Patch quickly and consistently. Attackers are exploiting newly disclosed vulnerabilities at speed.
- Reduce the attack surface. Move towards a zero trust approach where users only get access to the specific applications and resources they need.
- Think platform, not point products. Identity, endpoint, network and application security are all connected, so bringing them together improves visibility and reduces gaps.
Zero trust in a nutshell: never trust, always verify.
Enter WatchGuard FireCloud Total Access
FireCloud Total Access takes the security services from WatchGuard's Fireboxes and elevates them into the cloud, so whether the user is in the office, at home or travelling, the security policies follow them. That includes Gateway AntiVirus, APT Blocker, botnet detection, IPS, WebBlocker and application control.
It's an always-on VPN service that still lets users get to remote resources like file servers, a NAS or CCTV. It just does it very differently. There are three core parts:
- Connection Manager: the agent installed on machines, which handles authentication and routes traffic to the POPs.
- FireCloud POPs (points of presence): WatchGuard's global network, where zero trust policies are enforced and traffic is inspected.
- FireCloud Gateways: secure, fast connectivity between POPs and private resources.
Rather than granting access to the entire network, FireCloud flips this on its head with application-specific access. With a traditional VPN, users connect to the office and get the keys to the kingdom. FireCloud says you can only access these specific resources.
How it works
FireCloud sits inside WatchGuard Cloud, so if you already manage your firewalls there, the layout will look very familiar. Instead of firewall rules, there are access rules, defined by users and groups.
Authentication. Zero trust is all about identity. You can use AuthPoint, WatchGuard's own MFA solution, sync with your directory, hook into Entra or use third-party SAML providers.
Deploying the agent. The Connection Manager can be pushed out from WatchGuard Cloud to everybody or individual machines. Windows, macOS, iOS and Android are supported, which matters because mobile devices have been undervalued for a long time.
Private resources. If you already have a Firebox, you add a FireCloud gateway, select your firewall and save. No policies or rules to create. Then you add resources, such as a file server, and assign them to users or groups in the access rules.
Day to day. Logging in is typically a one-time thing. Switch the laptop off, put it in your bag, switch it on again on the train and FireCloud reconnects automatically. Blocked web categories and malware are stopped just as they would be behind the firewall, granted resources like the office NAS or a mapped drive work as normal, and anything that hasn't been granted is blocked. The disconnect option can also be hidden, so users are always connected.
Speed. FireCloud is based on WireGuard, with probably around four times the throughput of a traditional VPN. If your users work with large volumes of data, it's going to be like chalk and cheese.
Put simply, the old way was a one-time authentication and then access to everything. FireCloud is constantly authenticating the user, the device and what you're connecting to, over and over.
Common questions
Why do I need FireCloud if I already have a patched VPN and endpoint protection?
It's about resource-gated access and layered defence. The further away from the user you can push protection, the less chance of them clicking on or downloading something malicious.
Can FireCloud restrict access to Office 365?
Yes. When connected, you always break out from a specific range of egress IPs, which you can use in conditional access rules so 365 is only accessible from those locations. We see breaches into 365 all the time, so this is a quick way to reduce that risk.
What about when hybrid employees are in the office?
Typically you'd have a rule on the Firebox denying FireCloud, so you break out through the firewall as normal. As soon as you move away, FireCloud starts connecting again.
Where are the WatchGuard FireCloud POPs based?
There's one in the UK, along with POPs across Europe, the States and Asia, with around another 12 planned. You connect to the nearest one, and if it went down you'd connect to another.
Does it synchronise with ThreatSync?
Yes. All FireCloud detections now go into ThreatSync, the XDR part of WatchGuard Cloud, giving a single point of correlation alongside your other WatchGuard products.
