Program Design · Regulated Medical Device Manufacturing

What a Good GRC Program Looks Like

An FDA-regulated manufacturer with connected products, a plant floor, and a cloud back end — three regulatory worlds that most programs run as three disconnected bureaucracies. Architecture, evidence model, metrics, and the ten ways it fails.

The Measure of the Program

A GRC program is good if it shortens time-to-clearance and time-to-patch. Everything else is overhead wearing a compliance costume.

Policy count, control-mapping breadth, and framework logos are inputs. The program earns its budget in exactly two currencies: how fast a compliant product reaches market, and how fast a vulnerability in the field gets closed without a field action. If neither number improves, the program is a tax.

PremarketCybersecurity documentation is generated by engineering as a by-product of building, not assembled by regulatory in the final six weeks.
PostmarketA CVE lands, reachability is determined in days, and the disposition — VEX, patch, correction, report — follows a decision rule, not a debate.
Plant & EnterpriseProduction and QMS software, OT networks, and cloud infrastructure sit in one risk model with the product, not in a separate IT silo.

Sections below are collapsed. The regulatory surface, architecture, metrics and failure modes are the load-bearing parts; open the rest as needed.

1 · The Regulatory Surface

Scope first, because most programs are mis-scoped from day one. A digital-manufacturing medtech company is simultaneously three regulated entities: a device manufacturer (quality system, design controls, premarket submissions), a production operation (automated equipment, MES, historians, plant OT), and a digital business (cloud services, customer data, enterprise IT). Each has its own regulator, auditor, and failure mode.

The 2026 alignment matters here. QMSR took effect 2 February 2026, incorporating ISO 13485:2016 by reference into 21 CFR Part 820. FDA re-issued both the premarket cybersecurity guidance and the Computer Software Assurance guidance on 3 February 2026 to align terminology with it. A program still built around the old QSR vocabulary will produce evidence that no longer maps cleanly to how a current investigator or reviewer reads it.

DomainBinding / authoritativeWhat it actually demands of the program
Quality system21 CFR 820 (QMSR, eff. 2 Feb 2026) · ISO 13485:2016 by reference · Part 11 · Part 803 MDR · Part 806Design controls, DHF/DMR, CAPA, supplier controls, records integrity. Security work must land inside these objects or it has no regulatory standing.
Product cybersecurityFD&C Act §524B (eff. 29 Mar 2023) · FDA premarket cyber guidance, final 3 Feb 2026 (supersedes Jun 2025 and Sep 2023) · postmarket guidance 2016SPDF tied into the QMS, AAMI SW96-aligned security risk file, threat model, machine-readable SBOM, VEX, vulnerability management plan, CVD policy, demonstrated patchability, scoped penetration test.
Security riskISO 14971 (safety) · AAMI TIR57 / AAMI SW96 (security) · IEC 81001-5-1 · IEC 62304Two linked but distinct risk processes. Security risk is expressed in exploitability; safety risk in harm. The bridge between them is the single most-audited artifact.
Production & QMS software820.70(i) · CSA guidance (final Sep 2025, re-issued 3 Feb 2026) · GPSV (Sec. 6 superseded) · Part 11Risk-based assurance by intended use and process risk — not blanket validation. This is what makes patching plant systems possible on a sane cadence.
OT / plant floorIEC 62443 · NIST SP 800-82Zones and conduits, asset inventory, segmentation from enterprise, change control that coexists with production schedules.
Enterprise & cloudNIST CSF 2.0 · ISO/IEC 27001 · SOC 2 · HIPAA where PHI flows · state privacy lawCustomer assurance and the control backbone everything else inherits.
Ex-USEU MDR · MDCG 2019-16 · NIS2 · EU Cyber Resilience Act · UK, Health Canada, TGA, PMDALargely the same evidence, different packaging. Build once, express many — or you will build it three times.
AI/ML functionsPCCP guidance · NIST AI RMF · ISO/IEC 42001 · EU AI ActChange control for models, data lineage, drift monitoring, and a predetermined change plan so retraining isn't a new submission each time.
Scoping error that costs the most: treating "cyber device" narrowly. §524B reaches devices with software that connect to the internet — but the premarket guidance addresses devices with cybersecurity risk generally. Programs that scope to "connected products only" discover mid-submission that a USB-serviceable pump, a Bluetooth wearable, or a device whose only network path is the clinician's laptop all carry documentation expectations.

2 · Program Architecture

The organising principle: one control set, three operating domains, one evidence store. Most failures are architectural rather than technical — the same control gets implemented three times, evidenced three ways, and audited three times, and nobody can answer a simple cross-domain question like "which fielded products contain this component."

GOVERNANCE — SINGLE ACCOUNTABLE EXECUTIVE Risk appetite · policy set · exception authority · board & audit-committee reporting · budget CPSO / CISO PRODUCT SECURITY Threat modelling · SPDF gates Security risk file (SW96) SBOM generation & VEX Premarket documentation PSIRT · CVD · advisories Legacy & EOL fleet policy Owner: Engineering + Product Security MANUFACTURING & OT OT asset inventory · zones/conduits CSA for production & QMS software Part 11 records & audit trails Patch windows vs. production Supplier & contract-manufacturer Line-down / ransomware continuity Owner: Operations + Quality ENTERPRISE & CLOUD Identity · endpoint · network Cloud posture · IaC controls SOC 2 / ISO 27001 attestation PHI handling · privacy Third-party risk IR & breach notification Owner: IT + Security Operations SHARED SERVICES — BUILT ONCE, CONSUMED BY ALL THREE Unified control framework one control, many mappings Evidence store automated collection, versioned Risk register bridge security exploitability → safety harm CAPA & change control the only route to a closed finding If any of these four exists in triplicate, the program is already failing — the symptom is that no one can answer a cross-domain question.
The four shared services are what separates a program from three coexisting bureaucracies. They are also the hardest to fund, because none of them is visible to an auditor as a deliverable.
The unified control framework — one control, many mappingsWhy the mapping direction matters more than the mapping itself

The common failure is building a spreadsheet that maps NIST CSF to ISO 27001 to SOC 2 to 800-171 and calling it a control framework. That is a crosswalk, and it operates nothing.

A control framework has controls as the primary object: each has an owner, an implementation, a test procedure, an evidence artifact, and a frequency. Frameworks map to the control, never the reverse. Add a framework and you add mapping columns, not new controls — unless a genuine gap appears, which is the useful signal the structure produces.

FieldExample
ControlThird-party components are inventoried per build with provenance and support status.
OwnerRelease Engineering (accountable), Product Security (assuring)
ImplementationCycloneDX generated in CI on every build; fails the pipeline on unresolvable provenance
EvidenceSigned SBOM artifact + pipeline log, retained per DHF retention rule
TestQuarterly: sample 5 releases, confirm SBOM hash matches shipped binary
Maps to§524B(b)(3) · FDA premarket guidance SBOM element · 820.50 supplier controls · SOC 2 CC3.2 · ISO 27001 A.8.8 · CSF ID.AM-02 · 800-171 3.4.1

The test of whether you have one: add a new framework obligation and count how many new controls you write. In a healthy framework the answer is usually zero to three. If it is thirty, you are running crosswalks, not controls.

3 · Lifecycle Integration — The Load-Bearing Design

This is where good and bad programs actually separate. Security either lives inside design controls, with defined gates and defined evidence, or it lives beside them as a review that happens too late to change anything.

DESIGN CONTROLS (21 CFR 820.30 / ISO 13485 CL.7.3) Planning & Inputs Design & Development Verification Validation & Transfer Submission & Market G1 G2 G3 G4 G5 G1 · Security plan Cyber-device determination Security requirements asdesign inputs, not "non-functional" Regulatory pathway + SPDF tailoring G2 · Threat model STRIDE / attack trees, beforearchitecture freeze Security risk file opened SBOM generating in CI G3 · Verification Every security requirementtraced to a test result SAST/DAST/fuzz on interfaces Residual risk dispositioned G4 · Independence Pen test scoped to realexternal interfaces Patchability demonstrated Manufacturing transfer & provisioning G5 · Submission-ready Documentation assembled fromartifacts already produced Labelling / security disclosures CVD policy live before first ship POSTMARKET — THE LOOP THAT FEEDS BACK INTO DESIGN IntakeCVD · feeds · researcher ReachabilitySBOM → affected fleet Safety impact14971 risk re-assessment DispositionVEX · patch · 806 · 803 Advisory & deployCSAF · adoption tracking Field vulnerability classes become threat-model inputs on the next programme — or you ship the same defect again CAPA & DESIGN CHANGE CONTROL Every gate failure, audit finding, and field security event terminates here — with an owner, a due date, and effectiveness verification. No parallel "security backlog."
Five gates, each producing an artifact that a submission consumes unchanged. The dashed return path is the part almost every program omits: field vulnerability classes feeding the next program's threat model.
If the premarket cybersecurity package is assembled in the last six weeks rather than collected from work already done, the program has failed regardless of whether the submission clears.
Linking the two risk registersThe bridge from exploitability to patient harm — the most-audited artifact you own

Safety risk (ISO 14971) and security risk (AAMI SW96 / TIR57) use different currencies. Safety asks how likely a hazardous situation leads to harm. Security cannot use probability the same way — an adversary is not a random failure, and historical rates don't predict attacker behaviour. So security scores exploitability, then hands off to the safety process at a defined join.

SECURITY RISK — AAMI SW96 · scored on EXPLOITABILITY SAFETY RISK — ISO 14971 · scored on HARM Threatactor & capability VulnerabilityCVE or design flaw Exploit pathreachable? Compromisedevice state change Hazardous situationclinical exposure Harmto patient/user THE JOIN Security hands over a credible compromise scenario; safety assesses it as a hazardous situation using 14971 severity.

What auditors and reviewers probe: whether the join is real. Two registers, no traceability between them, is the most common finding. The test is simple — pick a controlled entry in the security risk file and ask which safety risk file entry it maps to, and who assessed severity. If the answer is a person rather than a record, it fails.

The scoring trap: raw CVSS as a security risk score. CVSS describes a vulnerability in the abstract; it says nothing about whether the component is reachable in your architecture, whether compensating controls exist, or what the clinical consequence is. Use CVSS as an input, layer reachability and exploitability, and let the safety process own severity.

4 · Postmarket — Where the Program Is Actually Tested

Premarket is a finite project with a deadline. Postmarket runs forever, on a fleet you no longer control, with components you did not write, against a clock set by whoever publishes the CVE. This is the capability that separates a real program from a documentation exercise.

New vulnerability disclosedCVE feed · researcher · internal test In any product SBOM?answerable in minutes, or the program has a gap here no Record & close · no action yes Reachable in our architecture?code path · interface exposure · compensating control not affected Publish VEX statementwith justification — silence reads as unaware yes Safety impact under 14971?re-run the risk assessment — do not skip to patching controlled risk Routine remediationnext release · advisory · SLA clock uncontrolled Regulatory decision pointquality + regulatory + security, on a defined clock Part 806correction / removal report Part 803MDR if harm criteria met Coordinated disclosureCSAF advisory · ISAC · CISA The 2016 postmarket guidance's practical effect: controlled residual risk with timely remediation generally avoids reporting; uncontrolled risk does not. That determination is a record, not a conversation.
Every branch is a written decision rule with an owner and a clock. Programs that debate this fresh each time produce inconsistent regulatory outcomes and cannot defend them later.
PSIRT and coordinated disclosure — the operational minimumWhat has to exist before a researcher emails you, not after
  • A published intake path that does not route to legal first. security.txt, a monitored address, and a named human. Researchers who cannot find a path go to the press or the ISAC.
  • An acknowledgement SLA — 48–72 hours, tracked. Non-response is the single most reliable trigger for uncoordinated public disclosure.
  • Fleet queryability. Given a component and version, which shipped products, which software versions, which customers, which are still supported. If this takes two weeks, everything downstream is already late.
  • A patch delivery mechanism you have actually exercised on the deployed fleet — not a design intention. §524B's patchability expectation is about demonstrated capability.
  • Advisory tooling that produces machine-readable output (CSAF) alongside the human-readable version. Hospital security teams increasingly ingest it directly.
  • A legacy and end-of-support policy published before you need it, with dates. Announcing end-of-support in the same advisory as an unpatchable critical vulnerability is how a technical problem becomes a customer-trust problem.
The pattern that damages companies most: a researcher reports responsibly, gets a legal reply, waits, publishes, and the resulting story is not about the vulnerability — it is about the response. The vulnerability was survivable. The response is what customers remember and what regulators note.
The manufacturing layer — the half of scope most programs dropCSA, Part 11, OT segmentation, and the patch-window problem

Product security gets attention because FDA asks about it in a submission. The plant floor gets attention after a ransomware event stops a line. The programs that handle this well treat production and QMS software as in-scope assets with the same control set and a different risk profile.

CSA is a security enabler, not just a quality efficiency. The historical pattern — treat every production and QMS system as requiring exhaustive validation on every change — makes security patching so expensive that systems stay unpatched for years, which is itself the larger risk. The risk-based approach (assurance proportional to intended use and process risk, leveraging vendor evidence and system-generated records rather than screenshot binders) is what makes a monthly patch cadence achievable on validated systems. Programs that have not re-baselined their validation approach against the current CSA guidance are usually carrying a large, invisible, self-inflicted vulnerability backlog.

Plant-floor issueWhat good looks like
Asset inventoryPassive discovery on OT segments; every PLC, HMI, historian, and test fixture in the same CMDB as enterprise assets, flagged by zone.
Segmentation62443 zones and conduits with documented data flows; enterprise compromise cannot reach production directly, and vendor remote access is brokered and recorded.
Patch cadencePre-agreed maintenance windows tied to the production calendar, with a risk-based exception route — not ad hoc negotiation per patch.
Part 11 recordsAudit trails, e-signature integrity, and retention treated as security controls, since record tampering is both a data-integrity and a security event.
ContinuityLine-down scenario exercised, including "we cannot trust the MES." Recovery time objective agreed with operations, not assumed.
Contract manufacturers820.50 supplier controls extended to security: right to audit, incident notification clock, SBOM obligations flowed down in contract.

5 · Metrics That Mean Something

Most GRC dashboards measure activity — policies published, training completed, controls mapped, vulnerabilities found. All of those go up when the program is busy and none of them go down when the program is failing. Useful metrics come in two classes, and both are needed: leading indicators that predict a submission or field outcome, and lagging indicators that confirm it.

TELLS YOU SOMETHING WILL GO WRONG → COSTLY TO GAME → HONEST BUT BACKWARD-LOOKING THE ONES TO REPORT UPWARD VANITY USEFUL, EASILY MASSAGED % releases with threat model before architecture freeze Days: CVE published → reachability determined Repeat audit findings % fielded fleet on a supported version Submission cycle-time added by security review AI letters citing cybersecurity deficiencies Field actions traceable to a security defect Mean time to remediate, by severity Control test pass rate Training completion Number of policies published Controls mapped across frameworks Vulnerabilities found "Vulnerabilities found" rewards scanning. "Vulnerabilities found that were reachable, and how long they stayed open" rewards fixing.
Position on the vertical axis is how expensive the metric is to game. Repeat audit findings sits high because there is no way to improve it except by actually fixing root causes.
100%Releases with a threat model completed before architecture freeze
<5 dCVE publication to reachability determination across the fielded portfolio
0Repeat findings from the prior internal or notified-body audit
≤48 hAcknowledgement to an external vulnerability reporter
>90%Fielded units on a supported software version at 12 months
0Additional-information letters citing cybersecurity deficiencies
The full metric set, with why each one is chosenLeading, lagging, and the anti-metrics to stop reporting

Leading — predicts the outcome while you can still change it

MetricTargetWhy this one
Threat model complete before architecture freeze100%After freeze, findings become change requests nobody funds. This single metric predicts most premarket friction.
Security requirements traced to verification evidence100%Traceability is what a reviewer and an investigator both actually check. Untraced requirements are aspirations.
SBOM regenerated per build, hash-matched to shipped binary100%A submission-time SBOM is stale the moment it ships and useless for postmarket triage.
Components with an active upstream maintainer>95%Unmaintained dependencies are unpatchable by definition — this is a leading indicator of a future field action.
CVE publication → reachability determination<5 daysMeasures whether the SBOM-to-fleet query actually works. Most programs discover here that it doesn't.
Fielded units capable of receiving a patch without a service visittrend upDetermines whether remediation is a software task or a logistics project.
Production/QMS systems on a current patch baseline>90%Directly reflects whether the CSA re-baseline actually happened.
Security CAPAs open past due date0The earliest visible sign that findings are being parked rather than closed.

Lagging — confirms it, too late to change it

MetricTargetWhy this one
Additional-information letters citing cybersecurity0Each one is roughly a quarter of schedule. The most expensive lagging metric in the set.
Submission cycle time attributable to security review<10 daysIf the program is working, security is not on the critical path.
Repeat audit or inspection findings0Non-repeat findings are normal. Repeats say the CAPA process does not work, which is a far larger finding than the original.
Field corrections traceable to a security defect0The outcome the whole program exists to prevent.
Uncoordinated public disclosures0Measures the CVD process end-to-end, including tone.
Customer security questionnaires requiring bespoke responsetrend downA proxy for whether the trust package answers questions before they're asked.

Anti-metrics — stop reporting these upward

  • Policies published. Measures writing, not operating.
  • Controls mapped across frameworks. Measures spreadsheet width.
  • Vulnerabilities found. Goes up when you buy a scanner; says nothing about exposure.
  • Vulnerabilities closed without a reachability determination — closure by fatigue looks identical to closure by fix.
  • Training completion percentage. Necessary, never sufficient, and it always reads 97%.
  • Average CVSS of open findings. An average of context-free scores is a context-free number.

6 · How These Programs Fail

Ten failure modes, roughly in order of how often they are the actual root cause. Most are structural, and most look like adequate programs from the outside until a submission or an event tests them.

1 — Two risk registers that never meetSecurity risk in a security tool, safety risk in the QMS, no traceability

Security keeps a vulnerability register scored on CVSS; quality keeps the 14971 risk file. Neither references the other. The program looks complete from either side and cannot answer the only question that matters — what could this vulnerability do to a patient.

How it surfaces: an audit or reviewer asks for traceability from a security control to the hazard it mitigates, and the answer is a subject-matter expert's recollection.

Fix: make the join an artifact with a controlled identifier, not a meeting. Every entry in the security risk file names its safety risk file counterpart or explicitly records why there is none.

2 — Security engaged at design freezeReviewing what can no longer be changed

Security is invited once the architecture is set, finds real problems, and every finding becomes a change request competing against launch date. Predictably, most are deferred with a risk acceptance, and the accepted risks accumulate across releases into a portfolio nobody has ever totalled.

How it surfaces: a risk-acceptance register that only grows; or an submission cycle where cybersecurity questions add months.

Fix: gate G1/G2 with security requirements as design inputs. A design input has regulatory standing; a security recommendation does not.

3 — SBOM as a submission artifact rather than a build artifactGenerated once, by hand, for the filing

Someone assembles a component list for the submission. It clears. Eighteen months later a critical CVE lands in a widely-used library and the question "are we affected" takes three weeks, because the only SBOM describes a build that no longer exists.

How it surfaces: the first serious postmarket event. Nothing before that reveals it.

Fix: SBOM generation in CI, machine-readable, versioned with the binary, retained alongside the DHF, and queryable across the fielded portfolio in one operation.

4 — The CAPA graveyardFindings routed into a process that cannot absorb them

Security findings are correctly routed into CAPA, which is then overwhelmed by volume it was never designed for. CAPAs age, extensions accumulate, effectiveness checks become perfunctory. The process meant to demonstrate control becomes the primary evidence of its absence.

How it surfaces: ageing CAPA metrics, repeat findings, and eventually an inspectional observation about the CAPA system itself — a much worse outcome than the original finding.

Fix: triage criteria that distinguish a systemic finding (CAPA) from routine remediation (tracked in engineering with an SLA). Not everything is a CAPA; and separately, nothing that is a CAPA gets a parallel informal backlog.

5 — Validation burden makes patching impossibleA quality practice producing a security outcome

Production and QMS systems are held to a validation approach where any change triggers extensive revalidation. Patching becomes prohibitively expensive, so it stops. Years later the plant runs unsupported operating systems on systems that cannot be taken down, and the risk register does not reflect it because nobody scored it.

How it surfaces: ransomware, or an OT assessment that finds decade-old software on the line.

Fix: re-baseline against the current CSA guidance — assurance proportional to intended use and process risk, leveraging vendor evidence and system-generated records instead of manual documentation. This is one of the few places where the compliance-and-security interests point the same direction, and it is routinely left on the table.

6 — OT and the plant floor left out of scopeThe program stops at the products and the laptops

Product security is funded because FDA asks. Enterprise IT is funded because customers ask. Nobody asks about the historian, so the historian is unowned, unpatched, flat-networked to the business, and running a vendor remote-access tool with a shared credential.

How it surfaces: a production outage, usually not targeted — commodity ransomware that reached the line through a flat network.

Fix: one asset inventory covering enterprise, OT, and test infrastructure; 62443 zones with documented conduits; and operations co-owning the risk with a named accountable leader.

7 — Legacy fleet with no support policyDevices that outlive their software support, in the field, with no plan

Devices are in clinical use for a decade or more. Software support horizons are rarely stated at launch. When a vulnerability lands on a platform that can no longer be patched, the company has no pre-agreed position, and the announcement of end-of-support arrives in the same communication as the vulnerability.

How it surfaces: customer escalation, sometimes an ISAC advisory, occasionally a field action driven by inability to remediate.

Fix: published support lifecycle at launch with dates; a defined process for compensating controls where patching is impossible; and end-of-support communication decoupled from vulnerability disclosure.

8 — Legal-first response to security researchersConverting a survivable technical issue into a reputational one

A researcher reports responsibly. The reply comes from counsel, references terms of service, and asks about testing authorisation. The researcher waits, then publishes. The resulting coverage is about the response, not the flaw.

How it surfaces: once, publicly, in a way that is permanently searchable.

Fix: a published CVD policy with safe-harbour language, technical first response within 48–72 hours, and legal in an advisory rather than a front-line role. Credit researchers in advisories; it costs nothing and changes the relationship.

9 — Framework sprawl and mapping theatreBreadth of coverage substituted for depth of operation

The program adds frameworks — CSF, 27001, SOC 2, 800-171, 62443, AI RMF — each with its own artefacts, each mapped to the others, none operated to the point of producing reliable evidence. Effort goes into the map, and the map is not the territory.

How it surfaces: an audit where controls exist on paper and evidence collection is a scramble; or a customer questionnaire that takes four weeks.

Fix: one control set with mapping columns. Adopting a new framework should add columns, not controls. Measure evidence automation percentage, not mapping coverage.

10 — Diffuse ownershipThree functions each reasonably believing another owns it

Quality owns the QMS, IT owns infrastructure, engineering owns the product. Product security sits in the seam. Every function is behaving reasonably, and the gap is invisible until something falls through it — typically a postmarket vulnerability where nobody owns the regulatory reporting determination.

How it surfaces: slowly, then all at once, during an incident where the first two hours are spent establishing who decides.

Fix: one accountable executive for product security with authority over the gates and the postmarket decision rule, and a documented decision-rights matrix. The title matters less than whether that person can stop a release.

7 · Maturity and Sequence

1 · REACTIVESubmission-driven.Evidence assembled late. 2 · DOCUMENTEDPolicies exist.Controls inconsistent.Two risk registers. 3 · INTEGRATEDGates in design controls.SBOM in CI.CVD live.Risk registers joined. 4 · MEASUREDEvidence automated.Leading metrics drivedecisions.Reachability in days. 5 · ANTICIPATORYField vuln classes feednext-gen threat models.Security shortenstime-to-market.Competitive, not compliant. Most regulated manufacturers sit at 2. The jump from 2 to 3 is organisational, not technical — and it is the only jump that changes outcomes. Stage 5 is where the program starts paying for itself.
Stage 2 is stable and comfortable and produces most of the failures above. Nothing about stage 2 looks broken until a submission or an incident tests it.
A defensible 180-day sequenceWhat to do first when inheriting a stage-2 program

Days 1–30 — Establish ground truth

  • Portfolio inventory: every product, its regulatory status, its software support state, whether a current SBOM exists.
  • Determine which products meet the cyber-device definition and which carry cybersecurity documentation expectations regardless.
  • Locate both risk registers and test whether any traceability exists between them.
  • Time-box the exercise: given a named component, how long to identify affected fielded products. Record the real answer.

Days 31–90 — Close the two gaps that block everything else

  • SBOM generation into CI for products in active development; a best-effort reconstruction for the top fielded revenue products.
  • Publish the CVD policy and stand up intake with an acknowledgement SLA. This is low-cost and removes the highest-variance risk.
  • Define the postmarket disposition decision rule (the tree above) and get quality, regulatory and legal to sign it while nothing is on fire.
  • Name the single accountable executive and document decision rights, including who can stop a release.

Days 91–180 — Make it structural

  • Insert G1–G5 into the design control procedure so the gates are QMS objects with evidence requirements, not a security checklist.
  • Build the risk-register bridge and retrofit the two or three highest-risk products.
  • Re-baseline production/QMS software validation against the current CSA guidance to unblock patching.
  • Stand up the unified control framework with automated evidence for the ten highest-frequency controls.
  • Start reporting leading metrics to the board or audit committee — three of them, monthly, unchanged for a year.
Sequence matters: CVD and SBOM first, because they are cheap and they cap the worst-case outcome. Framework and control work later, because it is expensive and it caps nothing on its own.

Back to the Measure

Two numbers decide whether the program is real: how long a compliant product takes to reach market, and how long a reachable vulnerability stays open in the field.

Every artifact described above — the gates, the joined risk registers, the CI-generated SBOM, the disposition tree, the unified control set — exists to move one of those two numbers. Anything in the program that moves neither is overhead, and in a regulated manufacturer overhead is not neutral: it consumes the credibility the program needs on the day it asks engineering to stop a release.