Table of contents
At a glance
Which GMP evidence do we need before restarting after a cyberattack?
Establish the affected time period, data and processes, and how reliable the available evidence is. Check which version was actually restored and how subsequent transactions can be reconstructed or assessed. Coordinate technical containment, investigation and assessment of process impact. Base the return-to-use decision on data integrity, open findings and effective controls. System availability alone does not demonstrate correct or complete GMP records.
- Evaluate technical availability and professional reliability separately.
- Compare the restored status with required subsequent actions.
- Corrective actions may be initiated prior to the final determination of the root cause.
1. Document the GMP process alongside the technical incident
Source date: 7 October 2026. This article examines Human Medicinal Product GMP and the evaluation of affected computerized records. It is not a technical guide for attack prevention and does not comprehensively address data protection, security, or reporting obligations to authorities.
Annex 11 No. 13 requires reporting and evaluation of all incidents, not only system failures and data errors. For critical incidents, the root cause shall be identified; this forms the basis for corrective and preventive actions. Requirements relating to backup, access, and business continuity remain relevant for the resumption of operations following an attack.
The first technical question is not merely whether a particular server is affected. Determine which manufacturing, laboratory, warehousing, or quality decisions depend on its data. An indirect dependency via a report or interface may also be significant.
2. Containment and existing records protection
Immediate actions do not need to wait until the complete root cause is established. Consult with the relevant technical experts prior to implementing any interventions and document the impact these actions may have on available records. For example, a restoration procedure could overwrite subsequent local entries.
For GMP evaluation, time stamps, affected components, known data states, and decisions made during a deviation are important. Clearly distinguish verified observations, plausible assumptions, and open questions. A technical safety protocol may support part of the evidence, but it is not automatically equivalent to a complete technical change history.
Do not arbitrarily collect every protocol in the batch record. Ensure that information required for investigation and reconstruction is preserved in an appropriate format and clearly linked to the technical evaluation.
3. Which question is addressed by the given evidence?
A recertification report is more actionable when it links technical results to specific technical questions. The matrix below separates different validation tasks. It is not a fixed sequence that every company must follow.
| Question | Suitable evidence | What this evidence does not yet establish |
|---|---|---|
| Is the technical environment back in operational condition? | Documented restoration and functionality test | Completeness of all professional procedures |
| What data version does it contain? | Identified safeguards and reconciliation of included records | Evaluation of data generated later |
| Are values and decisions reliable? | Technical comparison of relevant data, histories, and independent records | Complete absence of any unknown technical issue |
| Are access points again controlled? | Evaluated accounts, roles, and relevant access controls | Accuracy of previously stored values |
| Can the process be reused? | Rationalized decision regarding scope, findings, and controls | Blanket release of other processes that have not been assessed |
4. Separate the security status from subsequent operations
In the following example, the intended starting point for the safeguarding includes the activities V01 through V08. Prior to the interruption, additional activities V09 and V10 were generated. An independent technical overview lists both follow-up activities. Following restoration, only activities V01 through V08 are available.
Two expected processes are missing. This does not imply that their values may be arbitrarily entered. Identify available original records, the current status of processing and decision-making, as well as potential downstream processing. Any re-creation must be controlled and must not inadvertently trigger duplicate processing.
If V10 is not reliably reconstructible, this limit remains part of the risk assessment. It cannot be closed by retrospective data entry or fabrication. Assess which decisions are excluded or possible under appropriate conditions.
| Activity | Available basis | Assessment requirement |
|---|---|---|
| V01–V08 | In the restored state, the following is included. | Identity, relevant values and required context verified |
| V09 | Original supporting document available independently | Processing status and controlled transfer including potential consequences to be clarified |
| V10 | Only an indication that the transaction was created is available | Content and status remain open; no fictitious reconstruction |
| Total inventory | Eight existing plus two anticipated future processes | Ten expected, initially eight available; completeness not yet verified |
5. Assess the implications for prior and open decisions
Consider not only the date of the next reactivation. The relevant period may begin earlier, for example if access data were misused prior to the visible failure. Its definition must be based on findings and justified uncertainty, not on an arbitrary standard window.
Check whether the affected data has already been used for laboratory assessments, manufacturing decisions, or batch certification. This may lead to further investigations and actions. However, there is no blanket equivalence between any cyber incident and a defective product test result or an automatically required recall.
A partial additional use can only be evaluated for a clearly defined and appropriately secured scope. An organizational control step is not effective merely because it is documented; what matters most is its actual coverage and practical implementation.
6. Retry with open issues transparently decided
Compile technical and professional assessments. Describe which data have been evaluated as reliable, which questions remain open, and which usage decisions are involved. A green technical test serves as evidence for this step, not as proof of all consequences of the incident.
Open issues must be addressed in accordance with their significance. Unclear, fundamental data integrity concerns may contradict the intended use. Less extensive residual tasks must not be conflated with them. Authority, deadline, and control mechanisms must align with the actual decision.
Subsequently, document whether the implemented actions have achieved the intended effect. The root cause investigation may yield further insights that could alter a previous assessment. Therefore, the closure of a reattempt does not automatically conclude the investigation of the incident.
7. Checklist for professional re-evaluation
This checklist adds GMP issues to the technical malfunction and security process. It should be associated with existing incident, deviation and continuity procedures so that no contradictory decisions are made.
- Affected systems, interfaces, time periods, and GMP processes documented.
- Verified results and pending assumptions documented separately.
- Appropriate documentation must be maintained for any changed actions.
- Source of restoration and actual data status identified.
- Evaluation of missing or questionable procedures including subsequent processing.
- Affected access rights and technical controls verified.
- Assessed the impact on previously made business decisions.
- Scope, limitations, and decision authority defined.
- Tracking open items and effectiveness of actions organized.
8. What you can prepare in advance before the next incident occurs
A SME does not require a newly established parallel organization in the event of an incident. Helpful are a current system description, known data owners, reachable operator contacts, and a clear mapping of critical processes. This enables IT to specifically explain what technically was restored.
For relevant systems, do not limit your testing to application startup; also verify access to technically meaningful records. Agree on how subsequent activities will be identified and evaluated if a backup does not contain the most recent working state.
Appropriate maintenance scheduling is determined by the process and its dependence on the system. Annex 11 does not specify a general revalidation cycle. A defined target must be aligned with technical capabilities and the actual planned alternate procedures.
Project experience and collaboration
Combine technical recovery with assessment of the affected data
After a security incident, we examine affected records and decisions alongside technical recovery. Which records were accessible, which changes can be identified, and which periods remain uncertain? A reliable backup may answer particular questions. Missing context or unexplained access needs further investigation. We help assemble the evidence for a justified return to use. The relevant specialists coordinate security analysis and any reporting decisions.
Frequently asked questions about Cyberattacks and data integrity
Can we use the application as soon as it becomes accessible again?
Availability alone is insufficient. Assess the restored data state, relevant functionalities, access controls, and outstanding findings. The decision must describe a specific, professional scope of intended use and take into account remaining questions.
Sources: [1]
Must the complete root cause be identified before action is taken?
Does a successful restore ensure data integrity?
Do we need to re-verify all historical data?
Can missing data entries be reconstructed from memory?
A reminder does not substitute for a reliable original record. Identify available bases and limitations. A controlled reconstruction must be traceable and must not invent or retroactively assign prior tests, approvals, or values.
Sources: [3]
Is a CAPA required after each cyber incident?
Critical incidents require root cause analysis as the basis for corrective and preventive actions, in accordance with Annex 11. The specific handling follows the significance and findings. Not every technical notification constitutes a similarly serious issue or demands the same level of action.
Do all systems have the same restart time?
No. Appropriate measures are based on the criticality of the process and the available alternatives. Define realistic requirements and assess the proposed procedures appropriately. Annex 11 does not specify a general time limit.
Sources: [1]
Can organizational controls provide temporary assistance?
Yes, provided that the specific risk is actually sufficiently understood and practically implementable. Describe the scope, responsible parties, and evidence of implementation. A mere instruction does not substitute for a lack of technical capability, if the process remains unreliable when carried out.
Must security protocols be fully included in every batch record?
What preparation is most beneficial for a small QA department?
Clarify data ownership, system dependencies, and reachable point of contact prior to implementation. Enhance recovery testing by including technically relevant data sets. This enables faster identification of recoverable decisions in an incident and identifies which records are still missing.
Sources: [1]
Sources and review date
Source date: 7 October 2026. Human medicinal products – GMP; Data integrity and professional reassessment following a safety-related incident. No comprehensive cyber defense or legal reporting guideline available. Tables and case studies are internal reference tools, not results of completed product testing.
- European Commission: EU GMP Annex 11, Computerised SystemsJanuary 2011 revision, in operation since 30 June 2011. Sections 7, 11–13 and 16 cover backups, periodic evaluation, security, incidents and business continuity.
- European Commission: EU GMP Chapter 1, Pharmaceutical Quality SystemRevision 2013; effective from 31.01.2013; §1.4(xiv): Root cause analysis, corrective and preventive actions, and effectiveness evaluation
- European Commission: EU GMP Chapter 4, DocumentationJanuary 2011; in effect since 30.06.2011; §§4.8–4.10: timely traceable records, correction and integrity
- EMA/ICH: Q9(R1), Quality Risk ManagementEMA/CHMP/ICH/24235/2006 Corr.2, published 31 January 2025. Q9(R1) has applied since 26 July 2023. Sections 4.3–4.6 and 5.1 cover risk assessment, uncertainty, control and review.
Sources checked: .
Combine technical revalidation and GMP assessment
We support you in structuring relevant data, evaluating existing evidence, and making a transparent and justified decision on usage. Technical security measures remain coordinated with the responsible experts.
- Computer system validation →
Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

