Request a quote
Emergency SectorsCareers About us Blog Get in touch
Meeting room with a wall covered in yellow sticky notes and two colleagues at a long table

An information security policy people actually read, and IT actually enforces

Most organisations have an information security policy, and almost nobody in them has read it. That is rarely the reader’s fault. It is a thirty-page document from a template, full of rules nobody checks, untouched since the day it was approved. Here is how to build one that works.

6 min read Written by the ITproposal team
Short answer

An information security policy works in three layers: a short document from leadership, topic rules for the people who follow them, and procedures for those who carry them out.

Every rule needs a setting or a check behind it. A rule that IT does not enforce and nobody measures is a wish, and an auditor will find it.

Write down what is already in place first, and list the gaps as actions with a date. Leadership approves the policy; a supplier cannot do that for you.

What an information security policy is, and what it is not

An information security policy states how your organisation handles information: what needs protecting, who is responsible, and which rules follow for everyone with access. It is not a list of technical controls and not a manual for the IT team. Those sit underneath it, not inside it.

There are three reasons you need one, and they all ask for the same document:

  • A standard. ISO 27001 requires a policy approved and communicated by top management. Which controls sit beneath it technically is covered in our ISO 27001 checklist; this article is about the policy itself.
  • Regulation. NIS2 asks the organisations in scope for policies on risk analysis and information system security, and the NCSC’s guidance in the UK points the same way.
  • A customer or insurer. A security questionnaire almost always opens with “do you have an approved information security policy?”, followed by when it was last reviewed.

None of the three asks for a thick document. They ask that it exists, that it is approved, that people know it and that you can show it is followed.

Three layers instead of one long document

The most common design is a single file holding everything, from leadership’s intent to the length of a password. Nobody reads it, because every reader needs a different part. Split it by reader instead:

  • The policy itself, one or two pages. For leadership, the auditor and the customer who asks. It covers why, the scope, who is accountable and how it is kept up to date.
  • Topic-specific rules. For every member of staff: an acceptable use policy for devices and email, passwords and MFA, personal devices, remote working, sharing information externally, reporting an incident. One topic per page, so a new starter reads what applies to them.
  • Procedures and work instructions. For those who carry them out: how an account is created and removed, how a backup is tested, who does what in an incident. This part often belongs to whoever runs your IT, and rightly so.

The benefit is not only readability. A technical procedure changes more often than a principle, and with this structure you can update a work instruction without leadership having to re-approve the whole policy.

What belongs in the top document

The policy itself can be short if it answers these questions:

  • Why. What is at stake for this organisation: client data, availability of the service, confidential designs. Two sentences about you, not a paragraph about everyone.
  • Scope. Which sites, systems and people it covers, including contractors and suppliers with access.
  • Roles. Who is ultimately accountable (leadership), who monitors it day to day, and who owns which information. A policy without a name against a role has no owner.
  • Principles. For example: access only on the basis of role, data is classified, incidents are always reported. The detail is in the rules underneath.
  • Exceptions. Who may approve a deviation, how it is recorded and how long it lasts.
  • Maintenance. When it is reviewed and by whom, and where the current version lives.

That fits on two pages. If it grows beyond that, it contains rules that belong a layer down.

Writing rules that are read and enforced

A rule works when staff understand it and someone can see whether it is followed. Three habits make the difference:

  • Write what someone does, not what they should feel. “Be alert to phishing” is advice. “Report a suspicious email with the button in Outlook and do not click the link” is a rule.
  • Add a short reason to every rule. A rule without a reason gets bypassed the moment it is inconvenient.
  • Use “must” only where you check. If laptops must be encrypted, you must be able to show that they are.

That last point is where IT comes in. Go through every rule with whoever runs your IT and note one of three things against each: a setting that enforces it (MFA required, disk encryption, no local admin rights), a check that measures it (a monthly report of devices missing updates), or the honest statement that it is an agreement resting on trust. That third category may exist, but it should be small.

A rule that cannot be met technically is either removed or added to the action list. A policy promising something the environment does not do is worse at audit than a policy that does not promise it.

The order of work: write down what is already there

Starting from an information security policy template means writing a policy for an organisation that does not exist. Start from your own environment:

  • Inventory what is already in place. MFA, backup, updates, the joiner and leaver process, who has admin rights. Whoever runs your IT knows this, and it is often more than you expect.
  • Write the rules as things work today. Then the policy is accurate on the day it is approved.
  • List the gaps separately, with an owner and a date. What is missing belongs on an action list, not as a promise in the rules.
  • Have leadership approve it, with a date and a version number, and agree when it comes back to the table.
  • Make it known. At induction, in a short briefing for the whole team, and with a record that each person has read the rules.

If your rules touch on monitoring staff, such as retaining logs or blocking websites, discuss that with HR first; employment law and data protection both have something to say about it.

Keeping it current, and the pitfalls we see most

An information security policy goes out of date quietly. Review it at least once a year and also at every major change: a new site, a move to the cloud, an incident, a new supplier with access. Record what changed at each review, even when the answer is “nothing”; that is the evidence you looked.

The pitfalls we come across most often:

  • A template with someone else’s name in it. Or with systems you do not have. An auditor spots that in minutes.
  • Rules without an owner. Nobody feels responsible, so nobody notices when they stop being true.
  • Exceptions agreed in the corridor. The director who does not want MFA, with nothing written down. Record every exception with an end date.
  • No measurement. Awareness you do not measure is an assumption. How to measure it is covered in our article on the security awareness report rate.

Where our work stops

What we do. We inventory what is already in place technically, check your rules against what your IT can enforce and measure, configure the settings that go with them, and supply the reports that show compliance. We also write procedures and work instructions for the part we manage. We are ISO 27001 certified ourselves, so we know the questions an audit asks about the policy.

What we do not do. We do not approve your information security policy or write it on behalf of leadership: the decisions about risk, scope and accountability are yours. We are not an auditor or a certification body, and we do not declare that your policy meets a standard or NIS2. Nor are we lawyers: whether a rule holds up under employment law, or which sanction goes with it, is a question for HR and a lawyer.

Frequently asked

Questions we get about this

The five that come up most often.

What should an information security policy include?

The purpose, the scope, who is accountable, the principles and how exceptions and reviews work. That fits on one or two pages. The specific rules for staff and the procedures for IT sit in separate documents underneath, so each reader only reads what applies to them.

What is the difference between an information security policy and an acceptable use policy?

The information security policy sets the direction and responsibilities for the whole organisation. An acceptable use policy is one of the topic rules beneath it, telling staff how they may use devices, email and the internet. You need both, and the second should refer to the first.

Who approves the information security policy?

Top management. They are accountable, and an auditor will ask for that approval with a date and version. An IT team or supplier can prepare it and provide the technical rules, but cannot approve it on leadership’s behalf.

How often should an information security policy be reviewed?

At least once a year, and also after major changes such as a new site, a move to the cloud or an incident. Record every review even when nothing changes; that shows the policy is maintained.

Can you write and sign off our policy?

We check the rules against what your IT can enforce, configure the settings and write the procedures for what we manage. Your leadership approves the policy itself, and we do not declare that it meets a standard; that is for an auditor. For the employment law side, go to HR or a lawyer.

Related services

Further reading

What the policy connects to.

Shall we hold your policy up against your environment?

Send us your current rules, or tell us which questionnaire arrived. For each rule we will tell you what your IT already enforces, what you can measure and what is still a promise.

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