DORA-Readiness

DORA

ICT business continuity under DORA: plan, backup and testing

Under the full framework, DORA requires an ICT business continuity policy, plans for critical or important functions, a business impact analysis and tests at least yearly (Article 11), plus backups with segregated restoration and recovery objectives per function (Article 12). Under the simplified framework of Article 16, an ICT business continuity plan approved by the management body brings these topics together, tested at least yearly. This page applies to both frameworks and shows a plan outline with references and a completed test record.

Tape library for backups in a dark room, rows of tape cartridges behind the glass door in warm interior light

Last updated: 2026-11-06

This page applies to both frameworks: to financial entities under the full framework of Articles 5 to 15 DORA and to entities under the simplified framework of Article 16, including, from 1 January 2027, the institutions under Section 1a(2a) of the German Banking Act (KWG) (here: KWG group) such as factoring and leasing firms. Which framework applies to you is covered on our page DORA applicability. This page covers the structure, backup and testing of business continuity, that is ICT contingency management or business continuity management (BCM) under DORA.

Diagram “ICT business continuity under DORA” with two columns. Left, full framework: ICT business continuity policy (Article 11(1)), business impact analysis (Article 11(5)), recovery objectives per function (Article 12(6)), business continuity plans and ICT response and recovery plans (Article 11(3) and (4)), backup with segregated restoration (Article 12), crisis management function (Article 11(7)), tests at least yearly with cyber-attack and switchover (Article 11(6)). Right, simplified framework under Article 16: business impact analysis approved by the management body (RTS Article 28(2)(d)(i)), one ICT business continuity plan with recovery levels and timeframes, backup and escalation plans (RTS Article 39), tests at least yearly for backup and restore (RTS Article 40). At the bottom a band: deficiencies found in tests go to the management body under both frameworks.
Tap the graphic to open it at full size.ICT business continuity under the full framework and the simplified framework. Sources: Regulation (EU) 2022/2554 Articles 11, 12, Delegated Regulation (EU) 2024/1774 Articles 24 to 26, 28, 39, 40, as of 6 November 2026, own illustration.

What does Article 11 DORA require?

An ICT business continuity policy that the management body approves, oversees and periodically reviews (Article 5(2)(e)), and plans based on it. The policy may be a dedicated policy or an integral part of the overall business continuity policy (Article 11(1)). The paragraphs at a glance:

Article 11ObligationMicroenterprisesRTS (EU) 2024/1774
(1)ICT business continuity policy based on the identification under Article 8appliesArticle 24 (content of the policy)
(2)arrangements, plans and procedures: continuity of critical or important functions, quick response, containment, estimate of damage, communication and reportingappliesArticle 24(1)(b)
(3)ICT response and recovery plans with independent internal audit reviewplans yes, audit review noArticle 26
(4)implement, maintain and test business continuity plans, especially for outsourced critical or important functionsappliesArticle 26(4)
(5)business impact analysis (BIA) as part of the overall business continuity policyappliesArticle 24(1)(a)(i), Article 25(1), Article 26(1)
(6)tests at least yearly and after substantive changes, testing of crisis communication plans, review with test results and audit recommendationstests yes, without the obligation for cyber-attack and switchover scenariosArticle 25
(7)crisis management functionnot required-
(8)records of activities before and during disruption events readily accessibleapplies-
(10)report estimated annual costs and losses of major incidents on requestnot required-

Paragraph 9 concerns only central securities depositories, paragraph 11 the guidelines of the European Supervisory Authorities on paragraph 10.

BIA and recovery objectives: where the plan comes from

From the business impact analysis (Article 11(5)) and the recovery time and recovery point objectives derived from it for each function (Article 12(6)). What both contain is shown with example values on our page ICT risk management. What matters for the plan: the policy states that critical or important functions must be running again within these objectives after a disruption (RTS Article 24(1)(b)(ii)). Which functions they are is set by your list of functions.

The BIA is also mandatory under the simplified framework: the management body approves it (RTS Article 28(2)(d)(i)), and the ICT business continuity plan sets planned recovery levels and timeframes (RTS Article 39(2)(d)).

Backup under Article 12: what a DORA backup must achieve

It must be able to restore systems and data with minimum downtime. Article 12 under the full framework:

Article 12Obligation
(1)backup policies and procedures with scope and minimum frequency based on the criticality of information or the confidentiality level of the data, plus documented restoration and recovery procedures
(2)backup systems whose activation does not jeopardise security, availability, authenticity, integrity or confidentiality, with periodic testing of backup and restoration
(3)restoration with own systems on ICT systems that are physically and logically segregated from the source system and protected from unauthorised access
(4)redundant ICT capacities, for microenterprises see FAQ
(6)recovery time and recovery point objectives per function, taking into account critical or important functions and market efficiency
(7)checks and reconciliations during recovery so that data is consistent across systems

Under the simplified framework, according to BaFin’s supervisory notice of 21 August 2025 (section 4.4), Article 12 does not apply, and a backup concept is no longer an explicit requirement. Backup itself remains mandatory: the plan sets scope and minimum frequency based on the criticality of the functions using the data (RTS Article 39(2)(g)), plus restoration and recovery measures (point (f)), tested at least yearly (RTS Article 40(1)). We recommend a segregated backup protected against access there too, because after a ransomware attack only a backup the attacker has not reached will help.

Which scenarios must plans and tests cover?

Under the full framework, a whole list. Under RTS Article 26(2), the ICT response and recovery plans take into account, among others, cyber-attacks, the failure or insolvency of a key ICT third-party service provider, the loss of premises or data centres, lack of staff, natural disasters, insider attacks and large-scale power outages. Testing uses severe but plausible scenarios that challenge the assumptions of the plans and show whether staff, service providers and systems respond adequately (RTS Article 25(2)). Entities other than microenterprises test cyber-attacks and switchovers to redundant capacity (Article 11(6)). Results are documented, and weaknesses are analysed, addressed and reported to the management body (RTS Article 25(5)).

Under the simplified framework, the RTS expressly names only the cyber-attack alongside severe business disruptions (RTS Article 39(1)). Testing takes place at least yearly for backup and restore or upon every major change of the plan (RTS Article 40(1)), and deficiencies go to the management body (paragraph 3). These tests are part of business continuity. The testing programme for ICT systems under Articles 24 and 25 is described on our page DORA resilience testing.

What is different under the Article 16 framework

The simplified framework has no separate ICT business continuity policy, no ICT response and recovery plans under Article 11(3), no test concept for IT contingency plans as in BAIT and VAIT (BaFin’s supervisory requirements for IT in banks and insurers), no crisis management function and no cost reporting under Article 11(10) as explicit requirements (supervisory notice of 21 August 2025, sections 5 to 5.3). Their content is brought together in an ICT business continuity plan under RTS Article 39, approved by the management body, with escalation plans for communication (point (i)). How this plan fits into the evidence for the whole framework is shown by the evidence matrix on our page DORA Article 16: the simplified framework.

MaRisk AT 7.3 and MaGo: one contingency management or two?

Sectoral requirements on contingency management continue to apply alongside DORA. For institutions under MaRisk (BaFin’s Minimum Requirements for Risk Management), AT 7.3 requires an emergency concept for critical or important functions. How it fits with DORA business continuity is shown on our page MaRisk and DORA. For insurers under Solvency II, chapter 14 of MaGo (BaFin’s minimum requirements for the system of governance of insurers) applies, see DORA for insurers.

Work product: outline of an ICT business continuity plan

Blackfort Technology’s own outline, not an official template. It assigns the references of both frameworks to each chapter. Entities under the full framework use it for policy and plans together, Article 16 entities for the single plan under RTS Article 39.

ChapterContentFull frameworkArticle 16 framework
1 Purpose and scopeobjectives, scope with exclusions, period covered, link to general contingency planningArticle 11(1), RTS Article 24(1)(a)(i) to (iii)RTS Article 39(1)
2 Organisationroles, deputies, escalation paths, resources, approval by the management bodyArticle 5(2)(e), RTS Article 24(1)(b)(i)RTS Article 39(2)(a), (c)
3 Functions and BIAcritical or important functions, impact over time, dependencies on processes, assets and service providersArticle 11(5)RTS Article 28(2)(d)(i), Article 39(1)
4 Recovery objectivesrecovery time and recovery point per function, system objectives derived from themArticle 12(6), RTS Article 24(1)(b)(ii)RTS Article 39(2)(d)
5 Scenariosscenario catalogue with reasons why a scenario is relevant or notRTS Article 26(2)RTS Article 39(1) (at least cyber-attack)
6 Activationtrigger criteria, who decides, deactivation and return to normal operationsRTS Article 24(1)(a)(iv), Article 26(1)(a)RTS Article 39(2)(e)
7 Measures per functionimmediate measures, emergency operation, recovery, partial recovery, alternativesArticle 11(2) and (3), RTS Article 26(1)(b) and (e), Article 26(3)RTS Article 39(2)(f), (h)
8 Backupscope, minimum frequency, segregated restoration, integrity checksArticle 12(1) to (3) and (7)RTS Article 39(2)(g)
9 Service providersdependencies, contacts, failure of a provider, contractual commitmentsArticle 11(4), RTS Article 26(4)RTS Article 39(2)(d), (f) with final sentence
10 Communicationinternal and external communication, escalation, crisis team, reporting to BaFinArticle 11(2)(e) and (7), Article 14, RTS Article 24(1)(b)(vi)RTS Article 39(2)(i)
11 Recordslog of decisions and activities during a disruptionArticle 11(8)no explicit requirement, recommended
12 Teststest plan, scenarios, records, deficiency report to the management bodyArticle 11(6), RTS Article 25RTS Article 40
13 Maintenanceupdates after incidents, tests, new threats and major changes, auditArticle 11(3) and (6)RTS Article 39(2)(j)

Under both frameworks the plan must be documented and readily accessible in an emergency (RTS Article 26(1)(d) and Article 39(2)(b) respectively), including when the network and file storage are down.

Example: test record of a switchover exercise

Fictitious example, not a client. Muster-Kreditbank AG is the CRR credit institution from the example on our page ICT risk management. A data centre service provider runs its core banking system including payments at two sites. The recovery objectives for payments come from the BIA shown there.

ExerciseSwitchover of payments to the data centre’s secondary site, with crisis communication
Date, duration21 April 2026, 08:00-13:00, switchback on 22 April 2026, 06:00-07:30
Basisbusiness continuity plan chapters 6, 7 and 10, Article 11(6), RTS Article 25
ScenarioA cyber-attack takes down the systems at the data centre’s primary site. Payments are switched over to the redundant capacity at the secondary site. Assumption to verify: the secondary site is fully connected without any action by the bank.
Objectives under the planpayments usable again after 4 hours at most, data loss of 15 minutes at most, customer information within 60 minutes of the decision to switch over
Sequence08:00 scenario triggered by the exercise lead
08:10 crisis team convenes
08:50 management board decides on the switchover
10:00 customer information on the website, 70 minutes after the decision
10:40 secondary site reports ready for operation
10:55 reconciliation of booking balances completed
11:20 payments released to the business unit
ResultPayments usable after 3 hours 20 minutes, last data state 12 minutes before the start of the scenario: objectives met. Outgoing payment files backed up for 25 minutes until the connection to the clearing service provider at the secondary site was enabled. Customer information 10 minutes late. Switchback the next day without findings.
Deficiencies1. Connection to the clearing service provider at the secondary site not enabled in advance, the plan’s assumption did not hold. 2. The person designated for the public and the media could not be reached, and the plan named no deputy. 3. The reconciliation under Article 12(7) checked only account balances, not the totals of the payment files.
MeasuresOn 1: set up the connection permanently and include it in the contract with the data centre as a test item, head of IT, by 30 June 2026. On 2: name a deputy and add them to the plan, management board, by 15 May 2026. On 3: add reconciliation of file totals to the recovery checklist, head of payments, by 31 May 2026.
Report to the management board28 April 2026, deficiencies and measures noted (RTS Article 25(5)), follow-up in the quarterly report on contingency management

An institution under the Article 16 framework can use the same record. The RTS does not expressly require a switchover exercise there. What is mandatory is the yearly test of backup and restore (RTS Article 40(1)).

Three typical gaps

GapCauseEffectRemedy
Restore tested only file by fileThe restoration test recovers individual files, never a whole system with application and database.Whether a function is back within its recovery time remains unproven. The test does not show what Article 12(2) and RTS Article 40(2) require.At least once a year, a full restoration with time measurement and data reconciliation for each critical or important function.
Backup in the same administrative domainBackup servers and storage are attached to the same directory service and the same administrator accounts as production.An attacker who takes over production also encrypts or deletes the backups. The segregation under Article 12(3) is missing.Physically and logically segregated backup, separate access, immutable copies, practise restoration on segregated systems.
Recovery objectives stricter than the provider’s commitmentsThe BIA sets 4 hours, the contract with the data centre commits to 24 hours.The plan cannot be met in an emergency, and nobody notices. The dependency under RTS Article 39(2)(d) or Article 11(4) is not secured.Compare the BIA objectives with the contractual values, renegotiate the gap or plan fallback procedures, and include the provider in the tests.

How we help

Institutions of the KWG group moving from BAIT on 1 January 2027 build their ICT business continuity plan under RTS Article 39 in the transition from BAIT to Article 16, including an exercise. Other institutions under the simplified framework, such as small investment firms, small IORPs or insurance holding companies, start with the Article 16 gap analysis. For entities under the full framework, the DORA gap analysis shows where policy, plans, backup and testing stand against Articles 11 and 12.

Frequently asked questions

Does DORA set fixed recovery times?+
Only for three groups. Central counterparties and central securities depositories provide for at most two hours for their critical functions, and trading venues resume trading within two hours or a slightly longer period, with data loss close to zero (RTS (EU) 2024/1774 Article 24(2) to (4)). All other financial entities set the objectives themselves, per function and based on their BIA (Article 12(6) DORA).
Do microenterprises need redundant ICT capacities?+
Not necessarily. Article 12(4) DORA requires redundant capacities from financial entities other than microenterprises. Microenterprises assess whether they need them based on their risk profile. It is best to record that assessment in writing. That is our recommendation, DORA does not expressly require it. Article 12(5) requires a secondary processing site only from central securities depositories.
Does our IT service provider have to take part in the contingency tests?+
The tests include, where applicable, the services of ICT third-party service providers, including scenarios of their failure or insolvency (RTS (EU) 2024/1774 Article 25(2)(b)). For services supporting critical or important functions, Article 30(3)(c) DORA requires the contract to oblige the provider to implement and test business contingency plans. It is best to set out its participation in your tests expressly in the contract, see Managing ICT third-party service providers.
What has to be recorded during a disruption?+
Under the full framework, financial entities ensure that records of activities before and during disruption events are readily accessible once business continuity or recovery plans are activated (Article 11(8) DORA). In practice this means a log with time, decision, decision-maker and measure that can be reached even without the disrupted systems. It is also the basis for reporting and the post-incident review.

Read more

Sources

This content provides general technical and organisational information and does not constitute legal advice (no legal services within the meaning of the German RDG).