During a serious technology disruption, the bank rarely has all the facts. Operations may know that payments are delayed, technology may be investigating the cause, and customer teams may be receiving conflicting reports. The compliance challenge is to make and record timely reporting decisions while that picture develops.

For entities within DORA's scope, Regulation (EU) 2022/2554 creates the incident framework. Delegated Regulation (EU) 2024/1772 supplies classification criteria and thresholds. The reporting rules in Regulation (EU) 2025/301 set the initial, intermediate and final reporting time limits. The operating approach below should be implemented against those texts and the competent authority's submission arrangements.

Keep the legal decision separate from technical severity

An internal “severity one” incident does not automatically establish the DORA classification, and a lower technical severity does not settle it either. Maintain a classification record that maps facts to the applicable regulatory criteria. Capture the affected service, entity, customers, transactions, geography, duration, data consequences and cost information as relevant.

Assign an accountable decision-maker and a deputy. Technology should supply technical facts; business owners should explain service and customer effects; compliance should support interpretation and reporting. The arrangement must work outside normal business hours. A reporting process that waits for the next scheduled committee meeting is poorly matched to a live incident.

Record the clocks without rewriting history

For the initial notification, the general rule is as early as possible and within four hours of classification as major, no later than 24 hours after awareness. If classification occurs later than 24 hours after awareness, Article 5(2) requires the initial notification within four hours of that classification. The intermediate report is generally due within 72 hours of the initial notification, and the final report within one month of the intermediate or latest updated intermediate report. Credit institutions cannot use the weekend and bank-holiday deferral for initial notifications or intermediate reports. Apply the full provisions and relevant authority procedures in the operating rulebook.

Create separate timestamp fields for detection, awareness, classification, submission and recovery. Record the evidence supporting each. Do not overwrite an earlier estimate when better information arrives; preserve the original and explain the revision. This produces a defensible chronology without pretending that the first assessment was complete.

Submit a coherent picture under uncertainty

Use one controlled incident record to populate reports. Label estimates and identify who owns the underlying calculation. Where transaction impacts are still being reconciled, state the measurement approach and its limitations. Avoid false precision: an unexplained exact number can be less useful than a clear estimate with a defined coverage boundary.

A small reporting team can coordinate the narrative while specialists continue recovery. Require checks for inconsistent entity names, service descriptions, time zones and impact totals. Technical root-cause hypotheses should remain hypotheses until supported. The report should distinguish service restoration from clearance of a backlog and from completion of the investigation.

Rehearse an incomplete-information scenario

In a hypothetical exercise, online payments recover after a short interruption, but a reconciliation process reveals that a subset remains unprocessed. The team initially counts only the outage window. The exercise should test whether it recognises the continuing customer impact, revisits classification and updates the relevant report.

Include a supplier that cannot immediately explain its failure and a reporting portal that is temporarily unavailable. Test the bank's authorised alternative submission route and evidence retention. The objective is to identify decisions that become blocked, rather than to reward teams for producing a polished simulation report.

Connect reporting to prevention

After the incident, compare the first assessment with the final evidence. Which data arrived too late? Which decision required authority that had not been delegated? Did different teams use incompatible definitions of “recovered”? Convert those observations into specific process improvements.

Retain the classification record, approvals, submitted versions and receipt evidence together. Link the incident to service maps, supplier dependencies and remediation actions. Reporting quality and recovery quality should reinforce each other: the same disciplined understanding of impact helps both the authority and the bank decide what must happen next.