Fairness of AI-based applications in the financial sector

Technical rule [CURRENT]

DIN SPEC 91512:2025-09

Fairness of AI-based applications in the financial sector

German title
Fairness von KI-Anwendungen im Finanzsektor
Publication date
2025-09
Original language
German
Pages
35
Procedure
PAS

Free of charge

Format and language options

PDF download
  • Free of charge

Monitor with the Standards Ticker

This option is only available after login.
Easily subscribe: Save time and money now!

You can also subscribe to this document - together with other important standards in your industry. This makes your work easier and pays for itself after a short time.

Sparschwein_data
Subscription advantages
Sparschwein Vorteil 1_data

Important standards for your industry, regularly updated

Sparschwein Vorteil 2_data

Much cheaper than buying individually

Sparschwein Vorteil 3_data

Useful functions: Filters, version comparison and more

Publication date
2025-09
Original language
German
Pages
35
Procedure
PAS
DOI
https://dx.doi.org/10.31030/3628800
Loading recommended items...

Quick delivery via download or delivery service

Buy securely with a credit card or pay upon receipt of invoice

All transactions are encrypted

Overview

This DIN SPEC has been prepared by a DIN SPEC consortium (temporary committee) in the course of the PAS procedure. The preparation and adoption of this DIN SPEC has been carried out by the authors named in the foreword. This document applies to AI applications in the context of financial services. This document specifies the following: - a process for identifying appropriate concepts and metrics of fairness (for AI applications) in the financial sector; - a specific understanding of fairness may vary from person to person or depending on the situation and therefore cannot be formulated in universal terms; - to this end a basic framework for application- and situation-specific measurements and justifications in the context of financial services in practice shall be proposed, including: - a justification for the use of concepts such as equal treatment, equal rights, and equality; - relevant aspects of good measurement and justification of all concepts used; - assessment of proportionality with regard to results and expenditure; - handling of measurement results (evaluation, explanation, communication, need for action); - acceptable framework conditions for unequal treatment; - a collection of methods for operationalizing the identified concepts (for example, via fairness measurements). Identification of the respective strengths, weaknesses, and limitations of the operationalization approaches; - exemplary illustration of the process using relevant use cases. Moreover, in the context of its use cases, this document is limited to: - classification systems (excluding scorers and generative systems), - group fairness measures (excluding individual fairness and alternative fairness measurements). This document does not describe processes for adapting AI-based systems in order to meet fairness requirements.

Content

ICS

35.240.40

DOI

https://dx.doi.org/10.31030/3628800

Cooperation at DIN

Also available in
Loading recommended items...