What the National Cyber Security Centre’s (NCSC) New CAF Core Profile Means for Incident Communications
When critical systems go offline, who keeps everyone connected? The new NCSC baseline requires a response plan that works during an attack and has been tested.
Nov 17, 2026
·Blog
·Secure Communications
Listen to this article
:quality(90):no_upscale())
On October 30, 2026, the UK National Cyber Security Centre (NCSC) published the CAF Core Profile. It sets out the minimum level of cybersecurity the NCSC expects from organizations that run essential services or support the UK's critical national infrastructure. It is built on the Cyber Assessment Framework 4.0 (CAF), which many UK regulators use for their assessments.
For each of the CAF’s 41 contributing outcomes, the profile sets the level an organization is expected to reach:
a response plan built around the attacks it is most likely to face (D1.a)
the people, tools, and resources to carry out that plan during an attack (D1.b)
regular exercises that show the plan works (D1.c)
a process for reviewing each incident and fixing what went wrong (D2)
Several other areas of the profile only need to be partly met. Incident response is held to the full standard.
Four days after publishing the CAF Core Profile, the NCSC explained why it is raising expectations. Its annual report warns that UK critical infrastructure should be braced for a ransomware attack involving AI-powered malware. Launching the report, NCSC chief executive Richard Horne said the UK faces "a real prospect of an unprecedented rise in the number of cyber incidents" in the coming months. He said AI is helping attackers find and exploit long-standing weaknesses faster than organizations have ever had to contend with. He also warned leaders that they will have to answer to customers and employees when essential services stop.
More attacks mean more organizations will have to put their response plans into practice, not just test them in tabletop exercises. The CAF Core Profile asks whether those plans will hold up. One weak point rarely gets written down: how the response team will talk to each other once the attack is under way.
Recent survey data suggests few organizations have planned for it. A study of more than 9,500 organizations across Europe, North America, and Asia, published by CYGNVS and Marsh in October 2026, found that only one in three had a separate, out-of-band environment for running an incident response. Two in three had never tested recovery under real incident conditions.
Without a separate channel, the team falls back on corporate email, video calls, and the office phone system. All of those tools sign in through the same company directory that attackers go after first. In a serious attack, the organization may also switch those systems off to stop the attack from spreading. Either way, the response team loses its means of communicating at the moment it needs it.
The Bill and CI Fortify
NCSC says the Core Profile will also support the Cyber Security and Resilience Bill, which is in its final stages in Parliament and will bring more organizations under UK cyber regulation. Regulators can also set expectations above the profile where the threat calls for it.
In July, the NCSC co-signed joint advice that goes further. CI Fortify, led by Australia’s ASD with CISA, the FBI, and Canada’s Cyber Centre, tells operators to prepare to isolate vital operational technology (OT) and the systems that enable it from all other networks and to keep running that way for an extended period. It links isolation directly to incident response, meaning the organization must still be able to execute its response when those connections are cut..
Where Response Plans Break Down
Response plans are usually written and tested while everything is working. The tabletop exercise runs in a meeting room with email available. The contact list sits in the HR system, and the incident call runs over a video platform. All of that may work perfectly well in an exercise but become unavailable or compromised during a real attack.
The first problem is who else is listening. If attackers control the company directory, they may be able to sign in to anything that trusts it, including the incident call. A response team that plans its next move over a compromised channel is briefing the attacker. This is already happening. A joint advisory the NCSC released in July 2025 reports that Scattered Spider operators search victims’ chat and email for discussion of their intrusion and join incident response calls.
The second problem is reach. Keeping essential services running often means switching to manual operation at sites, run by engineers and shift staff who do not sit at a desk. The plan has to reach those people, confirm that each one got the instruction, and show who is where. Serious incidents also involve outsiders such as the regulator, NCSC, and sector partners, who need a channel they can rely on. A mailing list on a server that has just been switched off cannot do any of this.
A Checklist of What Assessors Will Look for
The CAF is not a certification matrix and no product makes an organization "CAF compliant." Assessors want proof that the plan works under attack conditions. For incident communications, these are the items to have in place and the evidence to have ready, with the Core Profile outcome each one supports.
☐ A separate channel for the response team that does not sign in through the company directory or run on corporate email. Evidence: the channel is named in the plan, with who uses it and when (D1.a).
☐ Contact details and call-out lists stored outside the systems an attack could take down. Evidence: where the lists are kept and when they were last checked (D1.b).
☐ A way to alert staff at every site and confirm who has responded. Evidence: delivery and acknowledgement records from a recent test (D1.b).
☐ Agreed, secure ways to reach the regulator, NCSC, and key partners. Evidence: named contacts and the channel agreed with each (D1.b).
☐ Exercises run with corporate systems assumed to be down. Evidence: exercise reports showing the team worked from the separate channel (D1.c).
☐ Response team phones and laptops enrolled and managed before an incident, including field-staff devices. Evidence: device inventory and the policies applied (B2.b).
☐ Administrator accounts for the device-management console and the separate channel that do not sign in through the company directory. Evidence: a list of administrator accounts and the authentication method used for each (B2.c).
☐ Records of what was said and decided during incidents and exercises. Evidence: logs used in the post-incident review (D2).
Every outcome on this list is set at fully met in the Core Profile. The same arrangements also help with the resilience preparation outlined in the framework.
None of this finds an intruder or restores a control system. It keeps the people doing that work able to make decisions without the attacker listening in.
How BlackBerry Secure Communications Supports This
BlackBerry® Secure Communications gives response teams a way to communicate outside the systems under attack.
BlackBerry® SecuSUITE® provides encrypted calls, secure messaging, and group chats on standard iPhone and Android devices. It does not depend on company email, the company directory, or collaboration tools, so response teams can communicate without relying on the systems under attack. BlackBerry SecuSUITE holds NATO Restricted approval.
BlackBerry® AtHoc® sends alerts to staff by app, text message, voice call, email, and desktop pop-up. It tracks who received each message and lets people reply, so the response team can see who is safe, who is on site, and who has not answered.
BlackBerry® UEM helps organizations configure and secure the devices response teams will use before an incident occurs. It applies security policy, keeps work data separate from personal apps, and can lock or wipe a lost or suspect device once an incident is declared.
To test how your incident communications would hold up when the corporate network cannot be trusted, book a resilience review with the BlackBerry Secure Communications UK team.
:quality(90):no_upscale())
:quality(90):no_upscale())