In short
According to Microsoft, the outage of 12-14 November 2026 was a technical problem, with no indication of malicious activity.
According to ACN, a NIS entity must notify if the outage exceeds the expected service levels defined under measure DE.CM-01 (incident IS-3), whatever the cause.
Supplier and customer may both have to notify: the supplier's obligations do not replace yours (ACN FAQ ISB.F.2).
The early warning must be sent within 24 hours of evidence that the threshold was exceeded; if it was not, document the assessment.
Impact analysis of SaaS services, alternative communication channels and supplier management are the basis of business continuity.
Between 12 and 14 November 2026 many companies worked without email, or with email coming and going, because of a global Microsoft 365 outage. According to updates from the official Microsoft 365 Status account, the problem started in Exchange Online, the service behind Outlook mailboxes, and then affected connectivity to Microsoft 365 services in several regions of the world. Microsoft stated that there was no indication of malicious activity or a cyberattack.
A few days later Italy's National Cybersecurity Agency (ACN) reminded NIS entities of a point many underestimate: even an outage caused by a supplier can be a significant incident to notify to CSIRT Italia. In this article you will find what Microsoft communicated, what ACN clarified, how to work out whether you must notify and what the event teaches about business continuity for anyone who relies on the cloud. This is not legal advice: for specific cases, speak to an expert adviser.
What happened to Microsoft 365 between 12 and 14 November 2026?
It was a technical infrastructure problem at Microsoft, not an attack. On 13 November Microsoft said it had no indication that the outage was the result of malicious or cyber activity. In the preceding hours it had explained that the elevated load was not driven by an increase in traffic, but by a number of servers becoming unexpectedly unavailable, which strained the remaining infrastructure.
The timeline according to Microsoft 365 Status
These are the main steps Microsoft communicated on its official account, with times converted to Italian time (CET). Recovery was gradual: for some hours access could work while mail delivery was still delayed, as Microsoft itself pointed out.
When (Italian time) | What Microsoft communicated |
|---|---|
12 November, 18:54 | Problems accessing Exchange Online mailboxes in the Americas |
12 November, 19:07 | The issue may affect Microsoft 365 service connectivity in any region |
12 November, 20:31 | The elevated load is due to unexpectedly unavailable servers, not more traffic |
12 November, 23:09 | A core process on the servers becomes overwhelmed, causing capacity loss |
13 November, 10:24 | The timeout for delayed email delivery is extended to 48 hours |
13 November, 16:39 | No indication of malicious or cyber activity |
13 November, 22:55 | Sustained recovery; extended monitoring, including the email backlog |
14 November, 7:09 | Service restored to the remaining affected customers and regions |
16 November | Preliminary post-incident report released in the admin centre |
Microsoft's preliminary report
On 16 November Microsoft made the preliminary post-incident report available to administrators in the Microsoft 365 admin center, and announced a final review with further details. If your company uses Microsoft 365, whoever looks after IT can read it there. For a customer company, though, the cause matters little: as we will see, what counts is the impact on your services.
What did ACN tell NIS entities about the Microsoft outages?
ACN clarified that a NIS entity must notify if, because of the outage, a significant incident occurs. It did so in a post on its official channels, which refers to ACN Determination no. 379907 of 19 December 2025 and to the frequently asked questions on incident notification published on the Agency's website.
ACN's message contains three practical points. First: for availability, the notification obligation is triggered if the outage exceeds the entity's own tolerance threshold, that is, the service levels documented when implementing measure DE.CM-01. Second: the underlying cause, the so-called root cause, is not decisive, and the entity must only assess whether one of the cases set out in the determination has occurred. Third: for supplier and customer relationships, the details are in FAQ ISB.F.2.
The Agency also urged all organisations to be more alert to calls, emails, links, chats and adverts that impersonate Microsoft staff to carry out malicious activity. A highly visible outage is an opportunity for fraudsters, and it is worth reminding employees straight away.
On causes, FAQ ISB.G.2 is explicit: the cause of the incident is not relevant when deciding whether it is significant. Incidents due to natural events, accidental malfunctions or human error are also covered, in line with the all-hazards approach of the NIS2 Directive.
Who must notify: the supplier, the customer or both?
Both may have to, each for its own services. ACN's FAQ ISB.F.2 covers a significant incident on the systems of a supplier that is a NIS entity providing services to a customer that is also a NIS entity. The notification obligation lies with the supplier, and also with the customer if the incident is significant in relation to the customer's own services and activities.
In practice the supplier handles its own obligations, but these do not replace yours. The fact that the outage originated outside your company does not exempt you from assessing its impact on your activities. For cloud services, FAQ ISB.F.3 states that when the incident occurs on the customer's systems the obligation lies with both customer and supplier, except for IaaS or hosting of the customer's infrastructure, where only the customer notifies.
The FAQs also point to a contractual obligation. When defining supply security requirements under measure GV.SC-01, the customer NIS entity must ensure that the supplier promptly reports security events that affect its services. That is a good reason to reread your contracts with critical cloud suppliers today.
What to check: the service levels of measure DE.CM-01
There is one thing to check: the expected service levels you have defined for your services and activities. Annexes 3 and 4 of Determination 379907 list the baseline significant incidents for important and essential entities, and the case linked to availability is IS-3.
Significant incident IS-3
IS-3 occurs when the NIS entity has evidence of a breach of the expected service levels of its services or activities, based on the expected service levels (SL) defined under measure DE.CM-01. Point 2 of the measure requires these levels to be defined and documented, including in order to detect significant incidents promptly. FAQ ISB.G.5 specifies that IS-3 mainly concerns availability.
If your service levels are documented, the question becomes concrete. For each service or activity hit by the Microsoft 365 outage, compare the actual duration and impact with the threshold you set. If the threshold was exceeded, an IS-3 significant incident has occurred and must be notified.
Your threshold counts, not the supplier's SLA
The reference point for IS-3 is the service levels of your own services and activities, not the availability the supplier promises in the contract. A cloud service SLA tells you what the supplier commits to; the DE.CM-01 threshold tells you how long one of your services can stop before the interruption becomes a significant incident.
That is why the two can diverge. An outage may stay within the supplier's SLA and still exceed your threshold if the affected service is critical to you. And as ACN pointed out, it does not matter that the cause is external: what is assessed is the effect on your services.
When the 24 hours start
The early warning must be sent to CSIRT Italia without undue delay and in any case within 24 hours of becoming aware of the incident. According to FAQ ISB.G.4, the deadline runs from when the entity has, after even a brief analysis, objective evidence that one of the cases has occurred. The full notification must follow within 72 hours of the early warning.
Also check whether the obligation already applies to you. According to FAQ ISB.G.3, for entities added to the NIS list in 2025 the deadline is nine months from the notice of inclusion; for those added for the first time in 2026 it runs from 1 January 2027. Voluntary notification remains possible in any case.
If the threshold was not exceeded, or if you are unsure
If the threshold was not exceeded you do not need to notify, but the assessment must still be carried out and documented. We recommend recording in the incident register the date, the services involved, the length of the outage and the comparison with the thresholds. If you are inspected, you will be able to show why you decided not to notify.
If you have not yet defined and documented the service levels for measure DE.CM-01, now is the time. Without written thresholds it is hard to tell within a few hours whether an outage is a significant incident, and the 24-hour deadline leaves no room for long debates. Also decide who carries out the assessment, who decides and who sends the early warning.
When in doubt, our advice is to notify. A precautionary notification is easy to manage, whereas failing to notify when it was required exposes you to consequences. ACN's FAQs also note that NIS entities can always voluntarily notify non-significant incidents, near misses and cyber threats.
What to do now: the post-outage checklist
The best time to act is now, while memories are fresh and people know what was missing. These are the actions to complete over the next few days, starting with the NIS assessment if you are a NIS entity.
Reconstruct which services and activities stopped, for how long and with what impact.
Compare duration and impact with the service levels documented under measure DE.CM-01.
If a threshold was exceeded, send the early warning to CSIRT Italia within 24 hours of the evidence.
If no threshold was exceeded, record the assessment and its reasons in the incident register.
Check that delayed email has actually arrived, especially orders, invoices and time-bound messages.
Remind employees to distrust calls, emails and chats presenting themselves as Microsoft support.
Reread contracts and SLAs with critical cloud suppliers, including the prompt-reporting obligation.
Record lessons learned with an owner and a deadline for each action.
If you have a certified management system, these records are also useful evidence for the management review and the audit, because they show that the organisation learns from events. For handling the first days after an incident, you will find a practical method in the guide on the first 72 hours of a data breach, which also helps you assess any data protection aspects with your DPO.
What does it teach about business continuity in the cloud?
It teaches that your continuity cannot depend solely on your supplier's. When a SaaS service stops, your processes stop: orders by email, customer confirmations, meetings, internal approvals. The supplier restores the service; organising work in the meantime is up to you.
Impact analysis of critical SaaS services
A business impact analysis measures what happens if an activity stops and for how long you can afford it. Applied to the cloud, it starts from processes: for sales, production, support and administration, ask which digital services are needed and what hidden dependencies exist, such as an ERP that sends order confirmations through company email or an application that uses the same Microsoft sign-in.
For each critical process, set the maximum acceptable downtime and the tolerable data loss, the classic RTO and RPO. They are also the natural basis for defining the service levels of measure DE.CM-01: a company with a sound BIA already has almost all the figures. You will find the full method in the guide to the business continuity plan for SMEs and in the approach to risks and opportunities in ISO standards.
Communicating when email and chat are down
You need an alternative channel agreed in advance, not improvised during the outage. If email and chat are on the same platform, the first problem is alerting colleagues, customers and suppliers. Choose an independent channel, such as phone calls, text messages or a tool from another company-approved supplier, and prepare an emergency contact list that can be consulted without logging into company systems.
Italian Legislative Decree 138/2024 provides for this for NIS entities: Article 24 includes among the measures, where appropriate, secured emergency communication systems. Decide in advance which channels are official, so employees know what to trust, and make sure IT knows where to read supplier updates: in Microsoft's case, the admin centre, the Microsoft 365 Status account and the public status page.
Cloud suppliers, SLAs and ISO standards: what to check?
Check that contracts, supplier assessments and your management system take into account the unavailability of a critical supplier. Article 24 of Legislative Decree 138/2024 requires NIS entities to adopt risk management measures that include business continuity, with backup management, disaster recovery and crisis management (letter c), and supply chain security (letter d).
Contracts, SLAs and supplier assessment
In the contract, check the guaranteed availability level, how it is calculated, the service credits, the channels through which the supplier communicates incidents and post-incident reports and, for NIS entities, the commitment to report promptly any events affecting your services. Remember that a service credit offsets part of the fee, not the downtime of your business.
A critical cloud supplier should be assessed and monitored like any critical supplier, recording the incidents that affected you and how they were handled. The assessment often leads not to a change of supplier but to compensating measures on your side. The method is described in the guide to supplier evaluation in ISO standards, and for cloud-specific standards see the guide to ISO/IEC 27017 and 27018.
ISO 22301 and ISO/IEC 27001
ISO 22301 is the international standard for business continuity management systems: it requires impact analysis, risk assessment, strategies, plans and procedures, and an exercise programme. For a cloud outage this means having a plan that covers the unavailability of a supplier and having tested it, for example with a tabletop exercise on a day without email. You will find the service on the ISO 22301 certification page.
In ISO/IEC 27001:2022, Annex A controls 5.19 to 5.23 cover information security in supplier relationships, from agreements and service monitoring to the use of cloud services; Annex A also includes controls on ICT continuity during disruption. The relationship between the two frameworks is explained in the guide on ISO 27001 and NIS2, and you will find the dedicated services on the ISO 27001 and NIS2 pages.
How Complaion helps you with NIS2, business continuity and information security
Complaion supports companies with information security and business continuity: NIS2 compliance, ISO 27001 certification and ISO 22301 certification, including as an integrated system. For NIS2 entities, we help you define the service levels and thresholds of measure DE.CM-01 and assess individual cases, as part of the wider NIS2 and ISO 27001 journey.
A dedicated Lead Auditor follows your company, analyses your starting point and proposes the most suitable path. The platform collects evidence, manages documents and procedures, keeps track of deadlines and prepares you for the audit. The Lead Auditor also carries out the internal audit, so you know whether your processes really work before external checks.
More than 1,000 clients have already chosen Complaion and, on certification journeys, our method has so far achieved a 100% success rate at the external audit. If you want to know where to start, contact us.
Frequently asked questions on the Microsoft 365 outage, NIS2 and business continuity
Was the November 2026 Microsoft 365 outage a cyberattack?
No, according to Microsoft. On 13 November the company said it had no indication of malicious or cyber activity. The preliminary report, released on 16 November in the admin centre, attributes the incident to a combination of technical factors; a final review has been announced.
Do I have to notify the outage to CSIRT Italia?
It depends on the impact on your services, if you are a NIS entity. According to ACN, the obligation is triggered if the outage exceeds the expected service levels you documented under measure DE.CM-01: in that case significant incident IS-3 occurs. The external cause does not matter. If the threshold was not exceeded, record the assessment anyway.
If the provider notifies the incident, do I have to notify it too?
Possibly. According to ACN's FAQ ISB.F.2, when the incident occurs on the systems of a supplier that is a NIS entity, the obligation lies with the supplier and also with the customer, if the incident is significant for the customer's own services and activities. The supplier's obligations do not replace your assessment.
When must the early warning be sent?
Within 24 hours of becoming aware of the incident. According to ACN, the deadline runs from when you have, after even a brief analysis, objective evidence that the threshold was exceeded. The full notification follows within 72 hours of the early warning. Also check from when the obligation applies to you.
Is the cloud supplier's SLA the threshold to use?
No. Case IS-3 refers to the expected service levels of your own services and activities, defined under measure DE.CM-01. The supplier's SLA is useful for assessing the supplier and for service credits, but an outage can meet it and still exceed your threshold.
How are BIA and service levels connected?
A business impact analysis establishes how long each process can stay down and how much data you can lose. These values are a natural basis for defining the expected service levels of measure DE.CM-01. An up-to-date BIA makes the assessment within 24 hours quicker.
Do I need ISO 22301 certification to manage these risks?
It is not mandatory. ISO 22301 is useful when you want a management system that keeps the continuity plan alive over time and lets you demonstrate it to customers and partners. If you already hold ISO 27001, you can integrate the two and share internal audits, reviews and documents.












