Skip to content

Commit 63aec4a

Browse files
authored
Merge pull request #323 from NGO-Algorithm-Audit/feature/structural_edits
Add empirical methods white paper in EN and NL
2 parents 1382bb6 + 85bd357 commit 63aec4a

File tree

8 files changed

+76
-3
lines changed

8 files changed

+76
-3
lines changed

content/.DS_Store

0 Bytes
Binary file not shown.

content/english/knowledge-platform/knowledge-base/Comparative_review_10_FRIAs.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
---
22
icon: fas fa-search
3-
featured: true
3+
featured: false
44
layout: article
55
type: knowledgebase_item
66
author: Algorithm Audit
Lines changed: 37 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,37 @@
1+
---
2+
intro: >-
3+
How empirical methods can help regulating AI
4+
featured: true
5+
layout: article
6+
type: knowledgebase_item
7+
facets:
8+
- value: type_white_paper
9+
label: white paper
10+
- value: subject_scientific
11+
label: scientific
12+
author: Algorithm Audit and Netherlands Bureau for Economic Policy Analysis
13+
summary: >-
14+
How empirical methods can help regulating AI
15+
weight: -16
16+
title: 'Empirical methods for supervising algorithmic profiling systems'
17+
subtitle: ''
18+
image: /images/knowledge_base/White_paper_Empirical_methods.png
19+
search:
20+
searchableText: tvt, toezicht, tijdschrift voor toezicht
21+
---
22+
23+
This article will be published in a special edition of the Journal for Supervision in Dutch *Praktische handvatten voor empirisch toezicht op profileringsalgoritmes*.
24+
25+
###### Abstract
26+
27+
When regulating algorithmic profiling systems, not only legal but also statistical information plays a key role.
28+
Using a Dutch public sector risk profiling algorithm as an example, we demonstrate that current frameworks,
29+
guidelines and soft law fall short in providing sufficient guidance for the interpretation of open norms within
30+
European non-discrimination law. We show that established methods from empirical science can help
31+
clarifying these norms. Building on the case-based example, we propose an assessment protocol designed
32+
to assist supervisory authorities in formulating targeted questions to examine indirect discrimination through
33+
algorithmic profiling systems used in the public and private sector. This approach builds upon existing
34+
legal frameworks, enabling supervisory authorities to effectively monitor algorithm-driven decision-making
35+
processes, even with limited resources.
36+
37+
{{< embed_pdf url="/pdf-files/knowledge-base/20250826_Empirical_methods_for_supervising_algorithmic_profiling_systems.pdf" width_mobile_pdf="12" width_desktop_pdf="6" >}}

content/english/knowledge-platform/knowledge-base/Public_standard_Meaningful_human_intervention.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
---
22
layout: article
3-
featured: true
3+
featured: false
44
type: knowledgebase_item
55
author: Algorithm Audit
66
summary: >-

content/english/knowledge-platform/knowledge-base/Public_standard_profiling.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
---
22
icon: fas fa-search
3-
featured: true
3+
featured: false
44
layout: article
55
type: knowledgebase_item
66
author: Algorithm Audit
Lines changed: 36 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,36 @@
1+
---
2+
intro: >-
3+
Praktische handvatten voor empirisch toezicht op profileringsalgoritmes
4+
featured: true
5+
layout: article
6+
type: knowledgebase_item
7+
facets:
8+
- value: type_white_paper
9+
label: white paper
10+
- value: subject_scientific
11+
label: wetenschappelijk
12+
author: Algorithm Audit en Centraal Planbureau
13+
summary: >-
14+
Hoe empirische methoden ondersteunen bij toezicht op AI
15+
weight: -16
16+
title: 'Praktische handvatten voor empirisch toezicht op profileringsalgoritmes'
17+
subtitle: ''
18+
image: /images/knowledge_base/White_paper_Empirical_methods.png
19+
search:
20+
searchableText: tvt, toezicht, tijdschrift voor toezicht
21+
---
22+
23+
Dit artikel is gepubliceerd in een thema-editie van het Tijdschrift voor Toezicht (TvT) als *Praktische handvatten voor empirisch toezicht op profileringsalgoritmes*.
24+
25+
###### Abstract
26+
27+
Bij toezicht op algoritmes en AI staat niet alleen het recht, maar ook de sta8s8ek centraal. Aan de hand
28+
van een concreet profileringsalgoritme lichten we toe dat bestaande kaders, handreikingen en so> law
29+
onvoldoende houvast bieden voor het invullen van open normen uit het gelijke behandelingsrecht. We
30+
illustreren dat beproefde methoden uit de empirische wetenschap handvaBen bieden om deze normen
31+
in te vullen. Aan de hand van het prak8jkvoorbeeld wordt een vragenlijst geïntroduceerd die
32+
toezichthouders ondersteunt in het stellen van gerichte vragen over indirect onderscheid door
33+
algoritmes en AI. Deze aanpak bouwt voort op weBelijke kaders en stelt toezichthouders in staat met
34+
beperkte middelen effec8ef toezicht te houden op algoritme-gedreven besluitvormingsprocessen.
35+
36+
{{< embed_pdf url="/pdf-files/knowledge-base/20250826_Empirical_methods_for_supervising_algorithmic_profiling_systems.pdf" width_mobile_pdf="12" width_desktop_pdf="6" >}}
254 KB
Loading

0 commit comments

Comments
 (0)