- UK GDPR Article 33 gives you 72 hours from becoming aware of a notifiable breach to report it to the ICO, not 72 hours from when the breach occurred
- Not every incident is notifiable. The threshold is whether the breach poses a risk to individuals' rights and freedoms
- A partial notification filed within 72 hours, with supplementary detail to follow, is treated far more favourably by the ICO than a complete report filed late
- Following a structured six-phase response (preparation, identification, containment, eradication, recovery, post-incident review) keeps you within the deadline and produces the evidence the ICO expects
- Maximum fines under UK GDPR reach £17.5 million or 4% of global annual turnover, whichever is greater, and late notification is itself a compliance failure separate from the underlying breach
Why the first hours define the outcome
Early containment limits how much data is exposed and how many systems are affected. A ransomware infection caught and isolated within an hour of detection typically affects a fraction of the systems that the same infection would reach after a full working day of undetected lateral movement. The same logic applies to unauthorised access: the sooner you revoke a compromised credential, the smaller the window in which it could have been used.
A documented, timestamped response also does something UK GDPR Article 33 explicitly asks for: it demonstrates that you had appropriate technical and organisational measures in place. The ICO's assessment of any breach weighs your posture and your response, not just the fact that a breach occurred.
What qualifies as a notifiable incident
UK GDPR Article 33 sets the threshold at risk to individuals' rights and freedoms, not at the existence of a security incident. A denial-of-service attack that knocks your public website offline for two hours, with no personal data exposed, is a security incident but is not notifiable. Unauthorised access to a customer database, a spreadsheet of applicant CVs emailed to the wrong recipient, or an unencrypted laptop containing employee records left on a train are all likely notifiable.
Assess risk against a small number of criteria: how sensitive is the data involved (financial details and health data carry more risk than a name and email address alone), how many individuals and records are affected, what could the data be used for if misused (identity theft and financial fraud represent higher risk than nuisance contact), and whether the exposure is reversible (a password can be reset; a leaked passport scan cannot be un-leaked).
When you are genuinely unsure whether a breach meets the notification threshold, the ICO's own guidance is to notify. A partial notification submitted within 72 hours, with a note that the investigation is ongoing and full details will follow, is treated far more favourably than a complete report filed after the deadline has passed.
The six phases of incident response
1. Preparation
Preparation happens before an incident, not during one. This means a documented response plan, a named team with defined roles, and playbooks for your most likely scenarios. Set clear escalation criteria for when the DPO, legal counsel, and executive leadership need to be brought in. See our incident response services for how we help organisations build this in advance.
2. Identification
Confirm the scope of what happened before you act on assumptions. Preserve logs and other evidence before you begin containment, since containment actions themselves (isolating a system, resetting credentials) can overwrite the evidence you need for both your ICO notification and any subsequent investigation.
3. Containment
Isolate affected systems to stop the incident spreading. Timestamp every action your team takes from this point. Containment is not the same as recovery: the goal here is to stop the bleeding, not to restore normal operations.
4. Eradication
Remove the root cause before you restore anything. Restoring a system to production before the vulnerability or malicious presence that caused the incident has been eliminated is one of the most common causes of reinfection.
5. Recovery
Restore from backups that have been verified clean, not simply the most recent backup available. Monitor restored systems closely for signs of reactivation before declaring the incident closed.
6. Post-incident review
Document what happened, what worked, and what did not. This record feeds directly into the accountability evidence UK GDPR expects you to maintain and should update your response plan and playbooks for next time.
Restoring a system before containing the incident is the most expensive mistake in incident response. It buys back a few hours of downtime and can cost weeks of reinfection and a much harder conversation with the ICO.
What to include in the notification to the ICO
Where a breach is likely to result in a high risk to individuals, UK GDPR Article 34 requires you to notify those individuals directly, without undue delay, in addition to your notification to the ICO.
How to build your plan before you need it
1. Define your team and emergency contacts
Name specific individuals, not job titles alone, for incident lead, technical response, legal, and communications. Keep contact details current and accessible outside your primary systems, in case those systems are the ones affected.
2. Document your notification criteria
Write down the risk factors your organisation uses to decide whether a breach is notifiable, so the decision does not have to be made from scratch under pressure.
3. Build playbooks for your top scenarios
Ransomware, unauthorised database access, a lost or stolen device, and a third-party or supplier data leak cover the large majority of incidents most organisations face. A specific playbook for each removes decision-making friction when time is short.
4. Establish an evidence preservation process
Define what gets logged, how, and for how long, before an incident happens. Evidence preserved incorrectly during the heat of containment is evidence you cannot use in your ICO notification or any later investigation.
5. Test the plan annually
A tabletop exercise, walking a realistic scenario through your plan with the actual people who would respond, reveals gaps that a document review never will.
6. Review the plan after every real incident
Treat every actual incident, however minor, as a live test of the plan and update it based on what you learn.
We help organisations build incident response plans that hold up under pressure, run tabletop exercises, and provide retained incident response support so the 72-hour clock is never a surprise. Get in touch to discuss your current plan.
This article provides general information and does not constitute legal advice. Whether a specific incident is notifiable depends on its facts and should be assessed with your Data Protection Officer or legal counsel.