The question no one asked before submitting the case chronology
When Sophie Chen at Halstead & Cole LLP pastes a case chronology brief into GPT-4o, she is doing three things simultaneously. She is using an AI system. She is processing personal data under UK GDPR. And she is potentially transmitting privileged communications to an external processor. Three obligations attach to that single action — and at baseline, the firm has documented none of them.
This is the data governance gap in its most concrete form. It is not an abstract compliance question. It is a question about a specific fee earner, a specific AI system, and a specific category of data — and what the firm's obligations are at the moment those three things intersect. ISO 42001 Clause 8.3 and UK GDPR require that question to have a documented, considered answer before the data enters the system, not after a client complaint or a regulator's enquiry.
Law firms face two simultaneous regulatory exposures when client data enters an AI system without a governance framework. First: SRA Principle 6 (client confidentiality) — the firm shall not disclose client information without authority. Submitting privileged data to an AI vendor's server may constitute disclosure, depending on the vendor's data terms. Second: UK GDPR Article 6 (lawful basis) — the firm shall identify a legal basis for every processing activity involving personal data. "We needed to draft the chronology" is not a lawful basis. Without a documented data governance register, the firm cannot demonstrate to the ICO or the SRA that either obligation was met — a dual-regulator exposure with enforcement consequences from both directions.
This article explains what Clause 8.3 requires, how it intersects with UK GDPR without duplicating it, the four data categories that law firm AI deployments typically handle, and how REG-AIMS-DATA-001 — the AI data governance register — closes Gap 6 across both frameworks simultaneously.
Two frameworks, one register: ISO 42001 Clause 8.3 and UK GDPR
The organisation shall establish and implement processes for the management of data used by or generated by AI systems, taking into account the intended use of the AI system and the interests and expectations of interested parties.
Data management processes shall address: data provenance and lineage; data collection; data preparation; data quality; data privacy and security; and data access and use.
Clause 8.3 is the data management clause of ISO 42001. It is deliberately technology-neutral — it applies whether the firm is using a cloud AI API, a self-hosted model, or an AI-enabled SaaS product. Its six components (provenance, collection, preparation, quality, privacy/security, and access/use) map directly onto the questions a law firm must answer about every data category it processes through an AI system.
Critically, Clause 8.3's "data privacy and security" component does not replace UK GDPR — it references it. ISO 42001 does not create new data protection obligations; it requires that existing obligations (under UK GDPR and the Data Protection Act 2018) be addressed within the AIMS framework. The result is that a law firm needs one integrated register that satisfies both instruments, rather than separate GDPR records and separate ISO 42001 documentation that are never cross-referenced.
- Documented data provenance for each AI system — where does the data come from?
- Data quality assessment — is this data fit for the AI system's intended use?
- Data access controls — who can submit data to each AI system?
- Data lineage — can the firm trace what data was used to produce a given output?
- Privacy and security controls — consistent with legal obligations
- Retention and deletion procedures for AI-processed data
- Lawful basis identified for each processing activity (Article 6)
- Special-category data: explicit consent or Schedule 1 DPA 2018 condition (Article 9)
- Data processing agreement with AI vendor if they are a data processor (Article 28)
- International transfer mechanism if data leaves the UK (Articles 44–49)
- Privacy notice updated to reflect AI processing if data subjects are clients
- DPIA (Data Protection Impact Assessment) for high-risk processing
The most available lawful bases for AI processing of client data in legal practice are Article 6(1)(b) — processing necessary for performance of the contract with the client — and Article 6(1)(f) — legitimate interests, where the firm's interest in using AI to provide efficient legal services is not overridden by the client's interests. Article 6(1)(a) (consent) is generally unsuitable as a primary basis for service delivery. For special-category data (health, race, religion, sexual orientation, criminal convictions), the firm requires a Schedule 1 DPA 2018 condition in addition to an Article 6 basis — most commonly paragraph 10 (legal claims) or paragraph 6 (health professional processing, where relevant). A firm that cannot identify a lawful basis for each data category it processes through an AI system is processing that data unlawfully.
Primary mapping — Information Security Management: ITIL 4's Information Security Management practice requires that information assets be classified, that access be controlled based on classification, and that the confidentiality, integrity, and availability of information be maintained across the service lifecycle. REG-AIMS-DATA-001 is the AI-specific expression of this practice — it classifies data by category (Special Category, Privileged/Confidential, Personal Data Standard, Anonymised), assigns access controls per classification, and documents the vendor security controls that govern each data flow. In ITIL 4 terms, each row of the register is an information asset control entry for a specific AI service.
Secondary mapping — Service Configuration Management: ISO 42001 Clause 8.3 requires that data lineage be traceable — the firm shall be able to identify what data was used to produce a given AI output. This is, in ITIL 4 terms, a configuration item relationship: the data input is a CI with a documented relationship to the AI system CI (registered in REG-AIMS-SYS-001) and to the output artefact. The data governance register extends the configuration model established in G3 into the data dimension — making REG-AIMS-DATA-001 and REG-AIMS-SYS-001 co-dependent instruments that shall be reviewed together at each quarterly AIMS cycle.
The SRA's confidentiality obligation and the ICO's AI guidance
The SRA requires firms to assess whether AI tool use is consistent with the duty of client confidentiality — including where client data is processed by third-party AI providers — and to document that assessment.SRA · Compliance Tips for Solicitors Regarding the Use of AI and Technology · 9 February 2026
The SRA's position is precise: it is not enough that the firm has a data processing agreement with the AI vendor. The firm shall actively consider whether use of that vendor is consistent with the duty of confidentiality, and that consideration shall be documented — otherwise it is not demonstrable. REG-AIMS-DATA-001 is the document that records that consideration for each AI system and each data category, satisfying the SRA's demonstrability requirement.
The ICO's AI and data protection risk toolkit requires that organisations processing personal data through AI systems document: the categories of personal data processed, the lawful basis, the retention period, any overseas transfers and their legal mechanism, and the risks and mitigations identified. This is a data governance register — and it is the same register that ISO 42001 Clause 8.3 requires. Law firms using AI tools face a single documentation obligation that satisfies both regulators simultaneously, and most have produced nothing for either.
The Pemberton Capital LLP questionnaire's fifth and final question — "how does your firm protect client data when using AI tools?" — is this gap in question form. A firm that can produce REG-AIMS-DATA-001 showing lawful bases, vendor DPAs, overseas transfer mechanisms, and data access controls for each registered AI system is answering Question 5 with evidence. A firm that says "we take data protection seriously" is not.
The four data categories and where law firm AI governance fails them
Law firm AI deployments typically involve four categories of data, each with distinct governance requirements. Most firms have no documented framework for any of them in the context of AI processing.
| Category | Examples in legal practice | ISO 42001 Cl. 8.3 obligation | UK GDPR additional requirement | Common failure |
|---|---|---|---|---|
| Special Category | Health data in personal injury / clinical negligence; immigration status; criminal convictions in criminal defence; religious beliefs in employment matters | Highest data quality and access controls; privacy impact addressed in impact assessment | Explicit consent or Schedule 1 DPA 2018 condition required in addition to Art. 6 basis; DPIA mandatory for systematic processing | Firm uses GPT-4o to draft medical chronology — no Schedule 1 condition documented; no DPIA conducted |
| Privileged / Confidential | Counsel's opinions; without-prejudice correspondence; client instructions; strategy documents | Data provenance documented; access restricted to authorised fee earners; vendor data terms reviewed for confidentiality | Not personal data if anonymised — but privilege is a separate legal concept not displaced by GDPR compliance | Privileged documents submitted to AI vendor without checking whether vendor's terms permit use for model training or logging |
| Personal Data — Standard | Client names, addresses, financial details; counterparty contact details; witness names | Data minimisation for AI use — submit only what the AI system needs for the specific task | Art. 6 lawful basis required; privacy notice must cover AI processing; DPA with vendor if they are a processor | Full client file submitted to AI for "summarisation" when only the relevant section was needed — no data minimisation applied |
| Matter Reference / Anonymised | Anonymised case facts; redacted documents; fictional scenarios for AI testing | Provenance documented even for anonymised data; re-identification risk assessed | GDPR does not apply to truly anonymous data — but pseudonymised data (with key held separately) remains personal data | Firm believes redacted documents are "anonymous" but counterparty names, case numbers, and dates remain — re-identification possible |
The data flow map — where Halstead & Cole LLP's data goes
The following diagram maps the data flows for each of Halstead & Cole LLP's four registered AI systems. Each row shows the data category entering the system, the vendor's processing location, and the governance status at baseline.
GPT-4o API →
Copilot Word →
Copilot Outlook →
LEAP AI →
exposure →
Building REG-AIMS-DATA-001: the integrated AI data governance register
REG-AIMS-DATA-001 is a single controlled document that satisfies ISO 42001 Clause 8.3 and the ICO's AI data governance requirements simultaneously. It is structured as a register of data flows — one entry per AI system per data category combination — rather than a policy document. Its purpose is to make the governance status of every data flow in the firm's AI estate visible, auditable, and actionable.
-
Map every data category to every AI system. Starting from REG-AIMS-SYS-001 Part A, identify every data category each system handles. The register uses four standard categories (Special Category, Privileged/Confidential, Personal Data Standard, and Matter Reference/Anonymised) but allows sub-categories where needed — for example, distinguishing health data from criminal conviction data within the Special Category column, as they have different Schedule 1 conditions.
-
Document the lawful basis for each data-system combination. For each row in the register, record the Article 6 lawful basis and, where special-category data is involved, the Article 9 condition and the Schedule 1 DPA 2018 paragraph. Where the lawful basis has not yet been determined, the row is marked "PENDING — system not cleared for this data category." A system is not cleared for a data category until a lawful basis has been documented and verified by COLP Priya Anand.
-
Confirm vendor DPA status for each system. For each AI system that acts as a data processor (i.e. processes personal data on the firm's behalf), confirm: whether a Data Processing Agreement is in place, what it covers, whether it explicitly addresses AI processing, and whether it includes a Sub-Processor Schedule. Record the DPA date and reference. For systems where the vendor is a data controller in their own right (rare in enterprise contexts but possible), record the basis for transfer and the privacy notice obligations that follow.
-
Record overseas transfer mechanisms. Where an AI system processes data outside the UK (OpenAI operates US servers; Microsoft operates across multiple regions), document the transfer mechanism — UK Adequacy Regulation, International Data Transfer Agreement (IDTA), or binding corporate rules. The zero-data-retention claim made by some vendors does not remove the transfer obligation — the data still crosses borders even if it is not retained.
-
Define data minimisation rules for each system. For each AI system, document what data categories fee earners may and may not submit. These rules are the operational expression of UK GDPR Article 5(1)(c) within the AI governance framework — and they are also the Clause 8.3 "data access and use" control. They shall be specific: "AIMS-SYS-001 — Personal data: submit only the specific passages relevant to the task. Do not submit full client files. Do not submit special-category data without prior COLP approval and a DPIA in place."
-
Flag systems requiring a DPIA and complete them. Any processing that is "likely to result in a high risk to individuals" requires a Data Protection Impact Assessment before the processing begins. In the AI context, this includes: systematic processing of special-category data; automated decision-making with legal or similarly significant effects; large-scale processing of personal data by AI systems. The register flags which system/data combinations require a DPIA; the DPIA itself is completed as a separate document and cross-referenced in the register.
OpenAI's enterprise API offers zero-data-retention as a contractual term — the firm's data is not used for model training and is not retained after the API session. This is a meaningful protection, but it does not eliminate the UK GDPR obligations. The data is still processed (transmitted, held in memory, used to generate output) — that constitutes processing under UK GDPR Article 4(2). The transfer still occurs. The lawful basis is still required. Zero-retention reduces the retention-related risk and the training-data risk; it does not displace the requirement for a lawful basis, a DPA, and an overseas transfer mechanism. REG-AIMS-DATA-001 documents all of these, not just the retention status.
REG-AIMS-DATA-001: AI Data Governance Register
The controlled document that closes Gap 6 is available below. It includes the full data governance register (one row per system per data category, pre-populated for Halstead & Cole LLP's four systems), the lawful basis and Schedule 1 condition reference table, the vendor DPA status log, the overseas transfer mechanism record, the data minimisation rules per system, and the DPIA trigger assessment checklist.
Full data governance register with lawful basis documentation · Vendor DPA status log · Overseas transfer mechanism record · Data minimisation rules per system · DPIA trigger assessment checklist · Special-category data clearance procedure · Version control and COLP sign-off column. Satisfies ISO 42001 Cl. 8.3 and ICO AI data governance requirements simultaneously.
Priority actions on issue: The two highest-priority rows at baseline are AIMS-SYS-001 (GPT-4o) processing personal data — the vendor DPA and overseas transfer mechanism shall be confirmed and documented within 14 days. The second priority is the blanket prohibition on special-category data in any AI system until a DPIA and Schedule 1 condition are in place — this prohibition shall be communicated to all fee earners in writing at the time REG-AIMS-DATA-001 is issued, cross-referenced to POL-AIMS-001.
REG-AIMS-DATA-001 is not a replacement for the firm's Record of Processing Activities (ROPA) required under UK GDPR Article 30. It is a supplement — an AI-specific data governance instrument that provides the detail the ROPA requires for AI processing activities. COLP Priya Anand or the Data Protection Lead shall ensure that the ROPA is updated to cross-reference REG-AIMS-DATA-001 for all AI-related processing activities, rather than treating them as separate documentation exercises. The two documents together constitute the complete UK GDPR Article 30 record for AI processing.
G6 in the final remediation stretch
With G6 complete, the firm has addressed all five of Pemberton Capital LLP's panel questionnaire questions with specific, documented instruments. The remaining gaps — G7 (client disclosure) and G8 (human oversight) — are the client-facing and operational controls that complete the governance architecture. G9 (supply chain) addresses Thornbury Legal Ltd.'s self-hosted model scenario.
The next article covers Gap 7: AI transparency and client disclosure under ISO 42001 Annex A.2. Once the data governance register identifies which processing activities involve client data, the disclosure procedure determines what clients need to be told — and when and how to tell them.
Quality gate record
The following gates were checked before publication. All 10 gates pass. This record satisfies ISO 9001:2015 Clause 7.5.3 (control of documented information) for the article series. Template: TMPL-AIMS-ART-001 v1.0.
- G1All regulatory claims carry named, citable sources. SRA instrument cited as Compliance Tips for Solicitors Regarding the Use of AI and Technology, 9 February 2026. ISO/IEC 42001:2023 Clause 8.3 cited directly. UK GDPR Art. 83(5) and DPA 2018 s.157 cited for fine quantum. Stat bar figures attributed to UK GDPR, UNUS London baseline assessments 2026, and scenario baseline. No fabricated citations.PASS
- G2ISO/IEC 42001:2023 Clause 8.3 correctly cited and characterised. Six data management components accurately listed. Prescriptive language — "shall" for requirements — correctly applied throughout. British English used in all clause callout text ("organisation" not "organization"). SRA obligation characterised as "requires" not "expects".PASS
- G3Scenario internally consistent with REF-AIMS-SCEN-001 v1.0. Firm: Halstead & Cole LLP. COLP: Priya Anand (Partner). Client: Pemberton Capital LLP, £620,000 per annum. Fee earner: Sophie Chen. AI systems: AIMS-SYS-001–004 correctly named. Second scenario firm: Thornbury Legal Ltd. No new named characters or firms introduced.PASS
- G4ISO/IEC Directives Part 2 prescriptive language applied correctly throughout. SRA confidentiality obligation characterised as "shall" in body text. Pull quote replaced with paraphrase and attribution — original text not reproduced verbatim. UK GDPR obligations characterised accurately as legal requirements. Data minimisation rule in Section 5 uses "shall" correctly.PASS
- G5REG-AIMS-DATA-001 download linked ungated in Section 6 doc-cta block. All cross-referenced document IDs present and correct: REG-AIMS-SYS-001, POL-AIMS-001, PROC-AIMS-HITL-001, DOC-AIMS-DISC-001, REG-AIMS-SUP-001.PASS
- G6Footer includes predecessor link (← G5 — AI Lifecycle) and successor link (G7 — Client Disclosure →). Section 7 documents all downstream dependency relationships (G7, G8, G9) with document IDs and ISO clause references. Breadcrumb updated to /legal/ path structure.PASS
- G7Standalone disclaimer section present between final-cta and footer. Includes: educational purpose notice, fictional scenario notice for Halstead & Cole LLP, Priya Anand, Sophie Chen, and Pemberton Capital LLP, document reference ART-AIMS-G006 v1.1, publication date, next review date, retention period, and canonical URL.PASS
- G8No policy or controlled document template content reproduced inline. REG-AIMS-DATA-001 is described and linked; its document-control block resides within the controlled document. Gate marked N/A — no inline template content requiring a document-control block.PASS
- G9H1 opens with primary keyword "AI data governance" within first 20 characters via title. Stat bar appears within first screenful. Schema.org Article JSON-LD with isPartOf series reference present in <head>. Canonical URL set: https://unuslondon.com/legal/iso-42001-gap-6-data-governance. Meta keywords present (6 terms including clause reference and "law firm").PASS
- G10IEC 82079-1:2012 notice hierarchy CSS comment block present above :root. .notice-danger class absent from stylesheet — removed entirely. DANGER not used anywhere. Section 1 notice reclassified from DANGER to WARNING (regulatory/dual-regulator exposure consequence). CAUTION used for vendor zero-retention commercial risk (Section 5). NOTE used for ROPA relationship (Section 6). ITIL 4 NOTE present with indigo styling, substantively mapped to Information Security Management and Service Configuration Management (Section 2). Eyebrow badge colour corrected from danger red to warning amber.PASS