# CAF Principle D1: Response and Recovery Planning — Cyber Security and Resilience Bill

> CAF principle D1 asks whether you have a risk-based incident response plan, the capability to carry it out and exercises that prove it works. For the Cyber Security and Resilience Bill (CSRB), it is where the 24-hour, 72-hour and customer notification duties become part of your response.

Canonical page: https://cybersecurityandresiliencebill.com/guides/caf-d1-response-and-recovery-planning
Updated: 2026-11-09

Status checked 13 November 2026: the Bill passed the House of Lords on 2 November 2026 and returns to the Commons, which is due to consider the Lords amendments on 24 November 2026. On 4 November 2026 the Commons extended its carry-over, so it lapses if not passed by the end of 29 April 2027. It is not yet law. References are to HL Bill 62 (as amended on Report), so wording and numbering may still change in the final Act. This guide is general information, not legal advice.

## Key points

- D1 has three outcomes: response plan, response and recovery capability, and testing and exercising.
- Your plan must be accessible even when your IT systems have been affected.
- Achieved capability needs duplicated key roles and arrangements for external incident responders.
- Build the Bill's 24-hour, 72-hour, CSIRT copy and customer notification duties into the plan, with named owners.

## What does CAF principle D1 cover?

Principle D1 asks whether you have an incident response plan that reflects your real risks, the people and resources to carry it out, and exercises that prove it works. It opens Objective D of the National Cyber Security Centre's Cyber Assessment Framework (CAF): minimising the impact of incidents. For organisations preparing for the Cyber Security and Resilience Bill (CSRB), D1 is where the Bill's reporting duties become operational. Your plan needs to say who notifies the regulator within 24 hours, who writes the 72-hour report, and who tells customers.

The CAF is a framework, not a law, but NIS regulators use it to judge whether security duties are met.

## What are the D1 contributing outcomes?

| Outcome | What it asks, in plain English | Rating |
|---|---|---|
| D1.a Response Plan | Is your incident response plan risk-based, complete, shared and available even when IT is down? | Not achieved / partially achieved / achieved |
| D1.b Response and Recovery Capability | Do you have the people, information, authority and backup arrangements to carry out the plan? | Achieved / not achieved |
| D1.c Testing and Exercising | Do you routinely exercise the whole response cycle with realistic scenarios and learn from it? | Achieved / not achieved |

## What does achieved look like for each outcome?

**D1.a Response Plan.** At *partially achieved*, the plan covers the systems behind your essential functions, comprehensively covers scenarios based on known, well-understood attacks, is documented, understood by your response staff, shared with stakeholders, reviewed regularly, and **accessible even when your IT has been affected**. *Achieved* means the plan rests on a clear understanding of your security risks. It covers the **complete incident lifecycle, roles and responsibilities, and reporting**, plans for previously unseen attacks as well as known ones, integrates with business continuity, supply chain response plans and supporting infrastructure such as power and cooling, and is understood by the business areas that run essential functions.

**D1.b Capability.** This is binary. To achieve it you need to know what resources and information a response will need, with arrangements to get them. Responders need the skills, knowledge and **authority** to act. Key roles must be **duplicated**, so no one person is critical. Backup mechanisms must let the essential function continue, possibly at a reduced level, if primary systems fail. And you need arrangements to bring in **external specialist incident responders** if needed.

**D1.c Testing and Exercising.** Also binary. Scenarios should draw on incidents at your own and other organisations or on threat intelligence, and be documented, reviewed and validated. Exercises should run routinely, test **every part of the response cycle** including restoring normal service, and feed findings into plans and protective security.

## Where do organisations usually fall short?

Paraphrasing the CAF's not-achieved indicators:

- **No documented plan**, or a plan that does not mention your essential functions.
- **A plan nobody knows.** Relevant staff do not understand it.
- **Missing resources.** No arrangements to get the people, tools or information needed in a real incident.
- **Responders without authority.** Teams cannot take containment decisions, such as isolating a system, without a long approval chain.
- **No fallback.** No way to keep the essential function going if primary systems are lost.
- **Partial exercises.** Testing only that backups work, or running exercises ad hoc, without feeding the results into lessons learned.

In our experience, a further gap is now common: **plans that predate the Bill** and say nothing about regulator notification, CSIRT copies or customer notices.

## What evidence could show D1 is achieved?

These are our suggestions, not NCSC requirements:

- An incident response plan with severity levels, roles, decision rights and a regulatory notification section naming who decides, who drafts and who approves
- Offline or out-of-band copies of the plan and contact lists, for example printed, or on separate devices or platforms
- A RACI showing deputies for every key role
- Retainers or framework contracts with incident response, forensics and legal support
- Playbooks for likely scenarios: ransomware, supplier compromise, cloud account takeover, OT disruption and data theft
- Exercise reports from tabletop and technical exercises covering detection to full recovery, with actions tracked to closure
- Evidence the plan was updated for the Bill's reporting duties and tested against the 24-hour and 72-hour deadlines

## A worked example

*This is our illustration, not an NCSC case study.* A relevant managed service provider runs a tabletop exercise in which ransomware spreads from its remote management platform to 40 customers. The exercise reveals that the plan is stored on the encrypted file server, that only the chief technology officer can authorise shutting the platform down, and that nobody knows who should notify the Information Commission.

Under D1, that points to gaps in D1.a (plan not accessible during an incident, no reporting section) and D1.b (single point of authority, no deputy). The fixes are simple but vital: keep an offline copy of the plan, give the on-call lead authority to isolate the platform, and name a primary and deputy owner for regulator notification, the CSIRT copy and customer notifications. Rerun the exercise against the Bill's deadlines to prove it.

## How does D1 relate to the Cyber Security and Resilience Bill?

Your plan should build in each of these duties from HL Bill 62:

| Duty | Timing | Where |
|---|---|---|
| Initial notification to your regulator: your name, the affected service and brief details | Within 24 hours of first awareness of a reportable incident | Regulations 11, 11A, 12A and 14E |
| Full notification: timing, duration, nature, impact including cross-border impact, any linked incident at another regulated organisation | Within 72 hours of the same moment | Same |
| Copy of each notification to the [CSIRT](/glossary/csirt) | At the same time as each notification | Same |
| Notify likely affected UK customers, explaining the incident and why they are affected (data centres, RDSPs, RMSPs) | As soon as reasonably practicable after the full notification | Regulations 11C, 12C and 14G |
| Customer updates on the incident and its impact | After the customer notification, as the provider considers necessary | Regulations 11C, 12C and 14G (added at Lords Report Stage) |
| Inform the public if the regulator directs you to, after consulting you (the regulator or CSIRT may also inform the public itself) | When directed | Regulations 11B, 12B and 14F |

Two practical points. First, notifications would have to be in writing and in the form and manner your regulator sets, so check regulator guidance when it is published. Second, under HL Bill 62 a mandatory incident must have had an **actual** adverse effect on the operation or security of your systems. Your plan should still log near misses for internal learning, as the CAF expects in D2. See the [24-hour notification runbook](/guides/24-hour-incident-notification-template), the [72-hour report template](/guides/72-hour-incident-report-template) and [customer notification](/guides/customer-notification).

## Where next?

Previous principle: [C2 Threat Hunting](/guides/caf-c2-threat-hunting). Next principle: [D2 Lessons Learned](/guides/caf-d2-lessons-learned). Recovery design is covered in [B5 Resilient Networks and Systems](/guides/caf-b5-resilient-networks-and-systems). For the bigger picture, read [CAF 4.0 explained](/guides/caf-4), the [CAF self-assessment guide](/guides/caf-self-assessment) and our [overview of the framework](/cyber-assessment-framework).

## Sources

- [NCSC Cyber Assessment Framework 4.0 (PDF)](https://www.ncsc.gov.uk/sites/default/files/documents/NCSC-Cyber-Assessment-Framework-4.0.pdf)
- [NCSC Cyber Assessment Framework collection](https://www.ncsc.gov.uk/collection/cyber-assessment-framework)
- [Cyber Security and Resilience (Network and Information Systems) Bill, HL Bill 62 (as amended on Report)](https://bills.parliament.uk/publications/68323/documents/8980)
- [Lords Report Stage debate, 26 October 2026 (Hansard)](https://hansard.parliament.uk/Lords/2026-10-26/debates/B413EFAE-E7D4-4742-986D-092884E0D8D1/CyberSecurityAndResilience(NetworkAndInformationSystems)Bill)

## Frequently asked questions

### What are the CAF D1 contributing outcomes?

D1.a Response Plan, D1.b Response and Recovery Capability and D1.c Testing and Exercising. D1.a can be partially achieved; D1.b and D1.c are rated only achieved or not achieved. The plan itself can be partially achieved, but having the capability to carry it out and testing it are all-or-nothing, so gaps there are particularly visible to an assessor.

### What should an incident response plan include for the Bill?

In our view: who decides an incident is reportable, who sends the 24-hour initial and 72-hour full notifications, who copies the CSIRT, who identifies and notifies affected customers, and who sends updates. Each role should have a deputy. Rehearse the notification steps in exercises.

### How often should incident response exercises run under the CAF?

The CAF says routinely rather than giving a frequency. Exercises should use realistic scenarios based on real incidents or threat intelligence, test every part of the response cycle including restoring service, and feed findings into your plans. Test against the Bill's deadlines.

### Does the CAF expect external incident response support?

For D1.b to be achieved, you should have arrangements to add external support, such as specialist cyber incident responders, if needed. A retainer or framework agreement agreed in advance is a common way to show this. Check the contract covers out-of-hours response.

