Self-assessment guide for artificial intelligence (AI) systems


The CNIL offers organisations an analysis grid through which to assess by themselves the maturity of their artificial intelligence systems with regard to the GDPR. It also describes best practices that can be followed.

Would you like to contribute? Write to ia[@]cnil.fr

IA - Guide d'auto-évaluation

Things to know before reading the guide

Introduction


Asking the right questions before using an artificial intelligence system

Fact sheet 1

Proportional integration of AI, specifying a clear objective

Collecting and qualifying training data

Fact sheet 2

Compliance with the GDPR when collecting and compiling a quality database.

Developing and training an algorithm

Fact sheet 3

Implementing best practices during this crucial phase.

Using an AI system in production

Fact sheet 4

Guarenteeing the quality and transparency of the system when in use.

Securing the processing

Fact sheet 5

Analysing risks and preventing flaws and attacks.

Ensuring individuals can fully exercise their rights

Fact sheet 6

Promoting transparency and rights for end-users.

Achieving compliance

Fact sheet 7

Assigning responsabilities and documenting the processing.

Other guides, tools and best practices

Further reading


Related articles

Euro numérique
01 February 2023
The CNIL is creating an Artificial Intelligence Department to strengthen its expertise on these ...
26 January 2023
Logo EDPB - CEPD
18 January 2023