Zum Hauptinhalt springen
ISO/IEC 27701:2025 · Annex A, Table A.1 · Control for PII controllers

A.1.4.4 Accuracy and quality

Updated on 5 min Reviewed by: Cenedril Editorial
ISO/IEC 27701:2025GDPR

An energy supplier moves its customer data to a new billing system. A mistake in the field mapping swaps date of birth and contract start for several thousand records. The automated credit check then treats long-standing customers as minors and demands advance payments. Corrections that the service centre makes in the old system never reach the new one. A.1.4.4 requires personal data to be kept as accurate, complete and current as its purpose demands throughout its entire life cycle, and this to be documented in a traceable way.

Wrong data does harm twice. It leads to wrong decisions about people, and it undermines trust in every analysis built on it. The more automated the processes, the further a single error travels.

What does the standard require?

The control combines a quality requirement with a duty to document.

  • Accurate, complete and current to the extent needed. The data has to serve the purpose it is processed for. That purpose also decides how precise it must be: errors matter most where they have consequences for people.
  • Prevent errors. Through policies, procedures or technical mechanisms, the organization keeps the number of inaccuracies as low as possible.
  • Respond to errors. For cases in which wrong data turns up anyway, there are defined procedures as well.
  • Documented and effective for life. These rules belong to the documented information, and technical system settings count as part of it. They apply from collection through to deletion. For the life cycle model, the standard refers to ISO/IEC 29101:2018, clause 6.2.

In practice

Prevent errors at the source. Validation rules for formats, required entries and plausibility, selection lists for recurring entries, address verification, duplicate detection. Each item is captured only once if possible and distributed from there to the other systems, so that there is one leading source.

Organize currency. Where data goes out of date, updates need a trigger: a customer portal in which people maintain their own details, returned mail as a signal, an annual confirmation for master data that carries weight, such as bank details.

Set up reporting channels for errors. Data subjects, employees and partners need to know where to report wrong details. The procedure sets out who checks, how quickly corrections happen, which systems are updated and when recipients of the data have to be informed.

Secure interfaces and migrations. Mapping tables are documented and tested. For migrations, reconciling record counts, field-level samples and an approval before go-live are standard.

Write technical rules down. A validation rule that exists only in the code is worthless for the audit. Record which rules apply in which system, and keep the description up to date through change management whenever something changes.

Typical audit evidence

For A.1.4.4, auditors typically expect:

  • Data quality policy or procedure with responsibilities for preventing and correcting errors
  • Description of technical validation rules and of interface mapping tables (→ Configuration & Change Management Policy)
  • Log of corrections with date, systems affected and recipients informed
  • Migration records with reconciled record counts and sample results (→ Change Log)
  • Results of regular quality checks, such as duplicate rates or returned mail figures

KPI

% of PII datasets with implemented accuracy verification procedures

Dataset here means a whole data collection, such as the customer base or the personnel files. Implemented means: validation rules are active at capture for that collection and a correction procedure is described. The target is 100% of the collections on which decisions about people are based.

Supporting KPIs:

  • Average handling time of rectification requests
  • Duplicate rate in the customer base
  • Share of corrections carried through to all affected systems

A.1.4.4 puts the accuracy principle of Article 5(1)(d) GDPR into practice. Data has to be accurate and, where necessary, kept up to date, and the controller takes reasonable steps to erase or rectify inaccurate data without delay, having regard to the purpose. Article 16 gives data subjects the right to rectification and to have incomplete data completed. While accuracy is contested, they can request restriction of processing under Article 18(1)(a). Article 19 requires recipients to be notified of rectifications (→ GDPR General Data Protection Regulation).

A.1.4.4 takes care of quality, while the other controls enforce the rights of data subjects:

Sources

Frequently asked questions

Do I have to update all data constantly?

No. The purpose is the yardstick. An invoice records the address at the time of invoicing and stays accurate even after the person moves. Data you act on today has to be current, such as delivery and contact details or information used for a credit decision.

What counts as documented information for this control?

Besides policies and work instructions, technical settings count too: validation rules in forms, mapping tables of interfaces, reconciliation jobs between systems. They have to be described so that someone else can find and check them.

How do I handle contested data?

Flag the item as contested and check it promptly. Until it is resolved, the data subject can ask under the GDPR for processing to be restricted. Once the error is confirmed, correct it in all systems and inform the recipients the data was sent to.

No Cookies!

This wiki collects nothing, bakes nothing and leaves nothing behind. There's nothing to consent to. Privacy doesn't get better than this.