Request a quote
Emergency SectorsCareers About us Blog Get in touch
Analyst wearing a headset in front of a wall of monitoring screens in a darkened room

DNS filtering: the safety net after the click

An email filter stops a lot but never everything, and someone clicking a phishing link is not an exception but a matter of time. DNS filtering works in the moment after: before the browser fetches the page, it asks where the page lives, and that is exactly the question you can refuse to answer.

6 min read Written by the ITproposal team
Short answer

DNS filtering blocks the lookup of known malicious domains before any connection is made. It works for every application and without decrypting traffic.

It only sees the domain name. A malicious file on a well-known cloud service gets through; that needs URL-level web filtering or endpoint protection.

Start with security categories, let the filter run in monitor mode for two weeks, and agree upfront who approves exceptions and what happens to the logs.

How DNS filtering works

Every click on a link starts with a question to a DNS server: at what address does this domain live? With DNS filtering that question goes to a resolver that keeps lists of domains used for malware, phishing or controlling infected devices. If the domain is on such a list, the device gets a block page or nothing at all instead of an address. No connection to the attacker’s server is ever made.

Three things make it attractive. It works for anything that uses DNS, not just the browser: an Office document trying to download something, or an infected device calling home. It decrypts nothing, so no certificates on devices and no view into the content of traffic. And it is quick to set up, often with a single setting on the router or firewall.

In the UK the NCSC runs a Protective DNS service for the public sector, and the same idea is widely available commercially for businesses. A useful extra in most products is blocking newly registered domains: phishing domains live briefly, and staff rarely need a three-day-old domain for work.

DNS filtering, web filtering or a proxy

The terms get mixed up, so first the definitions. A DNS filter sees only the domain name. URL-level web filtering (a proxy or secure web gateway) sees the full address and often the content, but has to break open encrypted traffic to do so, with certificates on every device and more to manage. A firewall with web categories sits in between and only covers traffic that passes that firewall.

Where the difference matters: attackers increasingly put their file on an ordinary cloud service, a shared folder or a forms service everyone uses. You cannot block that domain without blocking your own work. A DNS filter lets it through; URL-level filtering or the endpoint protection on the device has to catch it.

For most organisations the order is: DNS filtering everywhere first, then decide whether a heavier layer is needed. How that heavier layer comes together with remote access in a SASE design is covered in our piece on SASE and SSE.

Which categories to block

A DNS filter knows dozens of categories. Keep two kinds apart, because they call for different decisions:

  • Security categories: malware, phishing, command and control, cryptomining, newly registered domains, and lookalikes of your own name. Block these; nobody needs to be consulted.
  • Content categories: gambling, adult, social media, streaming, games. That is not security but policy, and the decision belongs to management and HR, not IT.

Run the filter in monitor mode for two weeks first. You see what would be blocked without anyone being stopped: a payment terminal using an unfamiliar domain, a supplier portal in an odd category. Then set up the exception process: who may release a domain, at whose request, and with an end date. An allow list without end dates becomes the hole the filter is routed around.

Where to put the filter

  • In the office: the router or firewall forwards all DNS queries to the filtering resolver and blocks DNS traffic to any other server. Without that second rule, any device can step around it with its own DNS setting.
  • The guest network: its own, stricter profile. It carries devices you do not manage.
  • Laptops at home and on the road: a small agent on the device or a setting pushed through device management, so the filter travels with the laptop. This is the part most often missing in a first rollout, and it is where most clicking happens.
  • Phones and tablets: through a profile from your MDM platform.

Do it in that order, and after each step check with a test domain that blocking actually works, in the office and on a laptop at home.

How it gets bypassed, and what to do about it

DNS filtering is a safety net, not a wall. The gaps are well known:

  • DNS over HTTPS in the browser. A browser sending its lookups encrypted to its own resolver skips your filter. Switch it off through browser policy or point it at your own resolver, and block the well-known public DoH services at the firewall.
  • VPN and proxy services. Block that category, and look at who tries: that is a conversation worth having.
  • A phone hotspot. That is why the filter belongs on the laptop itself and not only on the office network.
  • A hard-coded IP address in malware. There is no DNS lookup at all. That is what endpoint protection is for.

A filter that takes effort to get around is good enough. The target is the careless click, not the determined employee.

Logs, privacy and telling your staff

A DNS filter records which domains each device looks up, and therefore which sites staff visit. That is personal data under UK GDPR. Agree three things before switching it on:

  • Retention and access. Keep logs briefly, and only look at them during a security incident, by a small number of people.
  • Purpose. Security, not checking up on staff. Write it into your policy and tell your people.
  • Transparency. The ICO’s guidance on monitoring workers expects staff to be told what is monitored and why, and a data protection impact assessment where monitoring is likely to be high risk.

A filter with only security categories and short retention is easy to explain. One with content categories and long logs is not.

Where our work stops

What we do. We set up DNS filtering in the office, on the guest network and on laptops and phones through device management, enable the security categories, close the known bypasses and run the exception process. We run the filter in monitor mode first and go through the results with you before it blocks anything.

What we do not do. We do not decide which content your staff may visit: that is a decision for management, with HR. We do not read logs to check up on staff and we are not lawyers; whether your monitoring fits UK GDPR and employment law is for legal review. And DNS filtering does not replace email or endpoint protection: it is the layer in between.

Frequently asked

Questions we get about this

The five that come up most often.

What is DNS filtering?

DNS filtering blocks lookups of domain names known to be malicious, such as phishing and malware domains. The device gets no address back and never connects to the attacker’s server. It works for every application, not only the browser, and does not need to decrypt traffic.

What is the difference between DNS filtering and web filtering?

A DNS filter only sees the domain name being looked up. URL-level web filtering sees the full address and often the content, but has to break open encrypted traffic to do so. DNS filtering is simpler to deploy; web filtering also catches a malicious file on a well-known domain.

Is DNS filtering enough to stop phishing?

No, it is one layer. It stops many known and newly registered phishing domains, but not a page hosted on a trusted cloud service or an attack that uses no link. Combine it with email security, endpoint protection, multi-factor authentication and staff who feel able to report suspicious mail.

Can DNS filtering be bypassed?

Yes, for example with DNS over HTTPS in the browser, a VPN service or a phone hotspot. You limit that by setting browser policy, blocking public DNS services at the firewall and putting the filter on the laptop itself. The target is the careless click, not a determined user.

Can we see which websites our staff visit?

That is a legal and organisational question, not a technical one. DNS filter logs are personal data, and UK guidance expects staff to be told what is monitored and why. We use logs for security only; have your policy reviewed by a lawyer.

Related services

Further reading

The layers around a DNS filter.

Shall we run the filter in monitor mode first?

We put DNS filtering in monitor mode on your network and laptops for two weeks and show you what it would block. Then you decide what it is allowed to stop.

Practical IT knowledge in your inbox

New guides on management, security and the workplace, written by the people doing the work. No sales talk, and you can unsubscribe in one click.

We use your address for the newsletter only. See our privacy policy.

Request a quote Call