Illustration: an hourglass running fast beside a row of locked padlocks with one still open, for an article on patch management and the shorter time businesses have to apply security fixes.

Patch management: AI has shortened the clock

AI helps attackers find and use old software flaws faster. What patch management now needs in a UK business: the deadlines, the order, and who owns it.

Patch management is the routine of finding, testing and installing software security updates across every device, service and tool a business runs, against a deadline. AI has made the deadline matter more. Attackers now find and use old, known flaws faster, so a fix left for a month is a door left open for a month.

The UK’s National Cyber Security Centre made the point plainly when it launched its State of the Cyber Nation report on 3 November 2026. Its chief executive, Dr Richard Horne, warned of “a real prospect of an unprecedented rise in the number of cyber incidents”, and was clear about the cause: “not because AI has invented new ways to attack us”, but because “vulnerabilities that have been tolerated for far too long can now be discovered and exploited at a scale that was once unimaginable” (NCSC, 3 November 2026).

So the answer is the old discipline of keeping software up to date, run faster and checked properly.

What is patch management?

A patch is an update a software supplier releases to fix a fault, very often a security flaw. Managing patches means knowing what you run, knowing when a fix is released for it, installing the fix within a set time, and being able to show that it went on. It is one of the five controls we put in place for clients working towards Cyber Essentials.

The deadline already exists for any business that holds Cyber Essentials, the government-backed baseline. Its requirements say software must be updated within 14 days of release where the update fixes a flaw the supplier rates as critical or high risk, and the National Cyber Security Centre calls 14 days a reasonable period, adding that “any longer would constitute a serious security risk” (NCSC, Cyber Essentials requirements for IT infrastructure v3.3).

Most businesses are not there. The government’s Cyber Security Breaches Survey, published on 30 April 2026, found that 34% of UK businesses have a policy to apply software security updates within 14 days (DSIT, Cyber security breaches survey 2025/2026). The same survey found that 43% of businesses identified a breach or attack in the previous 12 months.

What has AI changed for patching?

It has changed how quickly old flaws are found, and how many fixes will need to go on.

In May the NCSC’s chief technology officer, Ollie Whitehouse, wrote that AI, in skilled hands, “is showing the ability to exploit this technical debt at scale and at pace”. The NCSC expects a “forced correction” across every kind of software, from commercial packages to cloud services, and has asked every organisation to prepare for a “patch wave”: a rush of updates across the whole technology stack as long-hidden flaws are disclosed, a number of them critical (NCSC, Preparing for a vulnerability patch wave, 1 May 2026).

For a business that patches once a month, by hand, when someone has time, that is the problem. The gap between a fix being published and an attacker using the flaw it fixes is the window that matters, and AI is closing it from the attacker’s side. That is the case for automating the routine now, before the wave arrives, which is how we set up the IT we run for clients.

What does good patching look like?

The NCSC’s guidance, read alongside the Cyber Essentials rules, comes down to six habits. We run them for our managed IT clients in this order:

  1. Know what you run. A list of every device, operating system, application and cloud service in use, including the ones a team signed up to on its own. You cannot patch what nobody knows is there.
  2. Start at the edge. Fix what faces the internet first (firewalls, remote access, email, websites), then cloud services, then the devices inside.
  3. Update by default. Turn on automatic updates wherever they exist, and make the NCSC’s “update by default” the written policy, with the exceptions named.
  4. Hold the 14-day line, and beat it under attack. Critical and high-risk fixes go on within 14 days. When a flaw is being actively exploited, especially on something internet-facing, the NCSC says to speed the process up.
  5. Replace what cannot be patched. Software out of support gets no fixes at all. Our article on Windows 10 reaching end of life covers the most common case.
  6. Check that it happened. A report that every device took the update, and an outside test now and then. Cyber Essentials Plus adds that test: an assessor checks a sample of devices, and a critical fix that has been available for more than 14 days is a fail.

AI shortens the time a business has to get these six right.

Do the AI tools you switch on need patching too?

Yes, and they need more than patching. Assistants, plug-ins and agents are software like any other: they have suppliers, versions and flaws, and the more they can reach, the more a flaw costs. The NCSC’s guidance on agentic AI asks organisations to match controls to how much autonomy an agent has, to log and monitor what it does, and to keep the ability to “pull the plug” (NCSC, Managing the cyber risk of agentic AI, 20 August 2026).

Few businesses have caught up. Of the businesses in the government’s survey that use AI or are adopting it, 24% said they had cyber security practices in place to manage the risks from it (DSIT, 30 April 2026). In practice that means the AI tools a team has switched on belong on the same list as the laptops: who owns each one, what it can reach, which version is running, and who is told when its supplier ships a fix. That is how we install AI for clients, with governance built to the ISO 42001 standard, because we have to live with what we recommend.

In-house or managed patch management?

Either can work. What fails is patching that belongs to nobody in particular. If you keep it in-house, name one person who owns the list and the 14-day clock, and give them the time to do it every week, not when a crisis allows.

If you hand it to a partner, ask three questions: what is on your list of what we run, how fast does a critical fix go on, and how do you show us it happened. A good answer to the third is an audited one. Our own information security is audited to ISO 27001 and our quality to ISO 9001, with zero non-conformances on both, and we hold Cyber Essentials Plus. We run managed IT and cyber security as part of the same company that builds and runs AI for clients, so the foundations under the AI are ours to keep sound.

Digital Planning is a board-level technology partner. We diagnose how a business runs, prove the change on the client’s own systems, then install, govern and run it. If you are not sure how long a critical fix takes to reach every device in your business, talk to us and we will find out with you.

Share this article:

LinkedIn

Author: Digital Planning

November 20, 2026

Patch management: AI has shortened the clock