Back online, still on hold: How GxP validation can delay cyber recovery in pharma and medtech

  • Blog
  • 6 minute read
  • October 15, 2026
Kareem Elwakil

Kareem Elwakil

Partner, Quality Transformation and Global CSV Practice Leader, PwC US

Nalneesh Gaur

Nalneesh Gaur

Principal, Cybersecurity, Data and Technology Risk, PwC US

Robbie Higgins

Robbie Higgins

Principal, Cybersecurity, PwC US

Key takeaways:

  • Cyberattacks moved deeper into life sciences supply chains in 2026. They disrupted ordering, shipping, production, and other operations across a growing number of companies.1
  • Getting systems back online doesn’t mean they’re ready for GxP use. For pharmaceutical, life sciences, and medtech companies, technical restoration may be only the first step. GxP systems also need the evidence and approvals required to return to GxP use.
  • Validation programs are designed for controlled changes with a defined scope. Cyber recovery is unplanned, can affect many systems at once, and introduces infrastructure or security changes with ramifications extending beyond the systems directly attacked.
  • GxP requirements extend the recovery timeline when they are not planned in advance. Even after systems are restored, production, product release, and shipment may be delayed while Quality, IT, and validation teams determine what changed and what assurance is needed.
  • Companies can prepare for GxP recovery before an incident. Establishing the assurance path for product release-gating systems moves scoping and evidence decisions off the critical path. Four actions make that possible.

Cyber recovery in a life sciences manufacturer has a second stage that many recovery plans don’t address. Although restoring systems returns them to technical availability, putting them back to work on GxP-regulated products requires documented evidence that the recovered environment remains fit for its intended use.

This distinction matters more as cyberattacks increasingly reach the medtech and pharmaceutical supply chain.

2026: The year cyber risk became an operations problem

Cyberattacks aren’t new to this sector, but the rate and concentration have escalated sharply. Attacks on healthcare providers plateaued through the first half of 2026 while attacks on the businesses that supply them, a group that includes pharmaceutical and medical device manufacturers, kept climbing.

Recent incidents across pharma, medtech, and the broader life sciences supply chain disrupted ordering, manufacturing, and shipping.1 The disclosed impact widened in the weeks after the first announcement as companies assessed the full scope, and systems came back before operations did, with order backlogs and manufacturing catch-up continuing long after restoration.

Bringing plants and systems back up is a series of recovery milestones rather than one event. For GxP-regulated manufacturers, assurance work adds another milestone: establishing that recovered systems can return to GxP use. When an organization has not established the assurance path in advance, that work lands on the critical path.

Restoration and GxP readiness are different milestones

Cyber recovery is planned against recovery time and recovery point objectives (RTO and RPO), which establish how fast a system returns to service and how much of its data can be recovered. Together, they define the technical recovery target.

GxP system readiness decisions, meanwhile, are made with different criteria: what changed during the incident and recovery, and what evidence is needed before the recovered environment can return to GxP use.

A technical restoration performed under an approved, pre-assessed recovery procedure may be supported by documented verification rather than a broader requalification effort. Recovery becomes more complicated when remediation changes system configuration, infrastructure, interfaces, data state, or other conditions on which validated operation depends. Those changes require an impact assessment, with assurance work scoped to risk, intended use, and the effect on product quality and patient safety.2 An assessment made at detection can be outdated by the time the full impact is known.

Evidence can also arrive slowly for legitimate reasons. A server may be held for forensic imaging before it can be rebuilt, and a database backup can be made available while its completeness and reconciliation still need assessment. Furthermore, the same infrastructure specialists are often split between containment, restoration, and evidence collection, straining the path to GxP readiness.

There’s also a disconnect between technical recovery and GxP use. IT recovery plans deliver a functioning environment, while Validation and Quality determine whether the recovered environment remains suitable for GxP use. Quality retains its approval and product-disposition responsibilities throughout, even when recovery work is led elsewhere. If IT, Validation and Quality haven’t agreed in advance on what evidence Quality needs, that question gets debated and settled during the incident. Production may be technically ready, but Quality can’t support release decisions.

Why cyber recovery breaks traditional validation assumptions

Traditional validation programs handle controlled change well: a defined system or software release, a known scope, an approved change process, and assurance activities sized against a documented impact assessment. Cyber recovery arrives in a different form. Scope is uncertain at the outset, multiple systems may be affected at once, and impact assessments that normally run sequentially must run in parallel. In some cases, the remediation itself becomes the largest source of change.

In practice, cyber remediation can tighten privileged access, endpoint management, and identity controls. Endpoint management became one prominent example in March 2026, when CISA urged US organizations to harden configurations after an attack that, according to public reporting, used compromised administrative access to wipe managed devices.3

Because those controls can sit underneath multiple GxP applications, a significant revision to a shared identity or endpoint-management layer widens the impact-assessment boundary to the systems that depend on it, even where application code hasn’t changed. A change to single sign-on rules, for instance, can bring configuration and installation qualification checks, along with targeted risk-based regression, into every application the assessment identifies as dependent on those settings.

During an outage, sites resort to approved paper-based recordkeeping. Those records create another issue after restoration when they must be reconciled, entered into electronic systems where required, verified against contemporaneous source records, and retained under applicable procedures. The resulting electronic records carry the same GxP controls for authenticity, integrity, accuracy, and retrievability as any other.4 That workload grows with the outage, and where business continuity procedures are either poorly established or unrehearsed, the administrative recovery can outlast technical restoration.

Current GxP and software-assurance frameworks do not prescribe a single post-incident requalification path. They require organizations to apply what their quality systems should already encompass, including risk assessment, change and configuration control, validation status, data integrity, and a documented rationale for the assurance scope selected. EU GMP Annex 11 requires controlled changes to computerized systems and periodic evaluation to confirm that systems remain in a valid state and compliant with GMP expectations.5

This is where cyber recovery strains validation decision-making, since the avoidable delay comes from determining the assurance scope while the recovery clock is already running.

Regulators are asking the same questions, with the July 2025 draft revision of Annex 11 making cybersecurity substantially more explicit and expanding requirements for system security, identity and access management, backup, and periodic review.6 The consultation closed in October 2025. As of September 2026, the 2011 Annex remains the published version in EudraLex.  

Four actions that shorten the recovery clock

  1. Establish the assurance path for release-gating systems before an incident. Work through the systems that stand between production and shipment and document the decision logic that Quality would use after a significant recovery event. Define which recovery conditions are addressed through documented verification, which drive targeted regression or qualification work, and which justify broader requalification. Record the preferred restoration sequence as well, since dependencies determine when Quality can begin its assessment. Document this within the disaster recovery and business continuity plans so your recovery team can use it.
  2. Connect disaster recovery and restore testing to the validation life cycle. Although restore testing is often designed primarily to prove that the environment and its data can be recovered, exercises for critical GxP systems should also generate artifacts Quality can reuse like approved protocols, expected results, configuration specs, reconciliation steps, and acceptance criteria. Tie that work to periodic evaluation so the recovery starts from an established procedure.
  3. Write recovery-evidence obligations into vendor contracts. Hosted and SaaS systems add a dependency the manufacturer can’t control directly. Configuration records, change histories, testing evidence, and restoration documentation come from the provider. If the agreement is silent on what will be supplied, by whom, and by when, that evidence arrives on the provider’s schedule rather than yours. Define the likely evidence requirements first, then incorporate them into vendor governance and renewals.
  4. Rehearse recovery jointly with IT and security through GxP use. Run the exercise beyond technical restoration to the point where a system is formally accepted into GxP use and product can move toward release. Rehearse the roles, responsibilities, and decision rights alongside the technical steps to be sure that governance and ownership gaps surface in the exercise rather than the incident. Include a scenario in which remediation changes shared identity or infrastructure controls, because that’s where the impact-assessment boundary expands fastest.

Don’t let the incident set your GxP recovery timeline

Scoping assurance work during an incident adds to the recovery timeline in three places—establishing what changed, completing the assurance work, and holding production or shipment decisions while the evidence is compiled. That last delay is the one customers feel. Your system can be technically restored and still not fit for GxP use. Until it is, product does not ship and nobody can give customers a date. Failure to revalidate before use is a compliance exposure in itself, and every record produced or batch released on that system can be challenged.

Establishing the assurance path in advance shortens the recovery timeline, limits the remaining work to what the incident truly warrants, and gives leadership a more defensible basis for estimating when regulated operations can resume.

This is a defined cross-functional project. Identify the release-gating systems, define the assurance path for each recovery scenario, align evidence requirements across Quality, IT, Security, and key vendors, and test the model through a joint recovery exercise. The result is a recovery plan that extends to GxP readiness rather than stopping at technical restoration, one that treats recovery as complete only when your organization can rely on a restored system for its intended GxP use and has the evidence to support that decision.

Our team works with life sciences and medical device manufacturers on validation strategy, quality systems, and cyber resilience, helping reduce the interval between technical restoration and return to GxP use by establishing the assurance paths, evidence requirements, and joint recovery exercises described here.  


  1. MedTech Dive, “Cyberattacks have plagued the medtech industry in 2026,” August 21, 2026.
  2. International Society for Pharmaceutical Engineering, GAMP 5: A Risk-Based Approach to Compliant GxP Computerized Systems, Second Edition, 2022; US Food and Drug Administration, Computer Software Assurance for Production and Quality Management System Software: Guidance for Industry and Food and Drug Administration Staff, February 2026, directed at software used in medical device production and quality management systems.
  3. Cybersecurity and Infrastructure Security Agency, “CISA Urges Endpoint Management System Hardening After Cyberattack Against US Organization,” alert, March 18, 2026.
  4. 21 CFR Part 11, including §11.10; European Commission, EudraLex Volume 4, Annex 11: Computerised Systems (2011), §§5 to 7, 9, and 17. §6 requires an additional accuracy check on critical data entered manually, by a second operator or by validated electronic means.
  5. European Commission, EudraLex Volume 4, Annex 11: Computerised Systems (2011), §§10–11, Change and Configuration Management and Periodic Evaluation.
  6. European Commission and PIC/S, draft revision of EudraLex Volume 4, Annex 11: Computerised Systems, published July 7, 2025; joint stakeholders’ consultation closed October 7, 2025. Draft text, not in force.

Cyber, Data and Tech Risk

Follow us

Required fields are marked with an asterisk(*)

Your personal information will be handled in accordance with our Privacy Statement. You can update your communication preferences at any time by clicking the unsubscribe link in a PwC email or by submitting a request as outlined in our Privacy Statement.

Hide