한국어

In September 2026 the Korea Laboratory Accreditation Scheme (KOLAS) accredited Pebblous as a data quality testing laboratory for ISO/IEC 5259-2:2024 and ISO/IEC 25024:2015. The AI Test & Evaluation Center carries out this testing and issues the reports.

A test report serves when data quality has to be evidenced by a third-party document, such as a public data construction delivery, a government R&D submission, or a procurement proposal. This hub gathers the guides on report types, the process, and the underlying standards. The application page is available in Korean.

Request a consultation Read the announcement

Accreditation details

Accreditation bodyKorea Laboratory Accreditation Scheme (KOLAS)
StatusAccredited testing laboratory
Accreditation no.KT1362
Valid throughSeptember 2030
Test standardsISO/IEC 5259-2:2024, ISO/IEC 25024:2015
Center locationDaejeon, Korea
Accredited testing siteSejong, Korea

Two types of test report

KOLAS-accredited test reportGeneral test report
StatusAccredited testing within the KOLAS scopeNon-accredited testing, outside the KOLAS scope
Test criteriaISO/IEC 5259-2, ISO/IEC 25024Criteria and methods set by the applicant
Test itemsStructured and unstructured AI training datasetsAI models and software, data processing systems
Test locationPebblous laboratory in SejongOn site at the applicant or remote, as agreed
Typical usePublic data project delivery, formal external submissionEvidence of project targets, internal verification

KOLAS accreditation recognizes the competence of a testing laboratory. It does not guarantee the quality of any product or dataset. A test report records measured results and is not a certificate.

Start here

The Pebblous AI Test & Evaluation Center announcement

Accreditation scope, the two report types, the application process, and what international acceptance does and does not mean.

ISO/IEC 5259 series hub

The international AI data quality standards behind accredited testing.

ISO/IEC 5259-2 quality measures summary

A quick reference to the quality characteristics and measures.

Image dataset quality guide

How the 5259-2 measures apply to image data.

Scoring ImageNet against ISO 5259

A worked evaluation of ImageNet using the measures in the standard.

Will Korean Data-Quality Reports Hold Up Under the EU AI Act?

How far a Korean test report goes in European regulatory procedures.

Related posts