Datatilsynet, January 2018
Norwegian Report on Artificial Intelligence and Privacy
On 2 February 2018, the Norwegian Data Protection Authority (DPA) published its Report "Artificial Intelligence and Privacy". This Report aims to describe and help us understand how our privacy is affected by the development and application of artificial intelligence (AI).
Target Group
The Report is written for people who work with, or who for other reasons are interested in, artificial intelligence. Ideally, engineers, social scientists, lawyers and other specialists will find this Report useful.
Content
After a Brief introduction on AI and data protection, the Report explores the tention between AI and privacy in four sections:
- Understanding AI (pp. 7-14):
First, the Report explains in detail how AI works by looking at machine learning, the value of training data and the danger of a resulting black box.
- Conflict with GDPR Principles (pp. 15-22):
Second, the Report examines which data protection principles enshrined in the GDPR are relevant for AI and scrutinizes the relationship between (1) algorithmic bias and the principles of fairness and non-discrimination, (2) AI and the principle of purpose limitation, (3) AI and the principle of data minimisation and (4) the black box of AI and the principle of transparent processing.
- Control Mechanisms (pp. 23-24):
Third, the Report takes a closer look at a DPA's supervisory competences with which the use of AI can be investigated and wonders how deep such investigation might go and how a "black box" might be investigated.
- Recommendations (pp. 25-29):
Fourth, the Report provides tools and solutions for data Controllers how to comply with data protection rules.
Datatilsynet, "AI and privacy", 2 February 2018
Datatilsynet, "Artificial intelligence and privacy", Report, January 2018