‘Technology must itself prevent human rights violations’

When deploying artificial intelligence, the government is bound by human rights and legislation. Municipalities are responsible for the assessment themselves, answers minister Bruins Slot (CDA) in response to parliamentary questions that were asked in response to the algorithm use of Rotterdam. New European legislation for the application of artificial intelligence should ensure that the developed systems take human rights into account.

Own responsibility

Member of Parliament Renske Leijten (SP) wanted to know which governments use algorithms and for what purpose, in response to the algorithms that Rotterdam uses to detect social security fraud. She asked whether the Minister of the Interior is prepared to investigate this. Municipalities themselves must ensure that they act according to the rules, Bruins Slot replies, and they must conduct their own investigations where necessary.

“In view of their own responsibility for compliance with the regulations, it is the responsibility of the provinces and municipalities to have an investigation carried out where necessary, via the Provincial Council or Provincial Executive and the municipal council or B&W,” writes the minister, who believes that there are A lot of research is already being done, also by independent bodies.


She believes that it is not up to the minister to judge Rotterdam’s use of algorithms, but she does tell what Rotterdam indicated: that the algorithm is efficient, that it would be unfeasible to study the large number of benefit recipients within one or two years. and that administrative burdens are avoided for some of the benefit recipients as a result of this working method. Also, the algorithm is only used for the selection and an employee always carries out the actual research.

AI Regulation

She also mentions State Secretary van Huffelen (Digitalization, D66), who recently argued for more digital regulations: ‘It is important that what applies in the offline world will also apply digitally’.

Bruins Slot elaborates on upcoming European legislation on artificial intelligence: the Artificial Intelligence Act, one of the measures by which Europe is trying to catch up with the digital backlog. The technology itself must prevent human rights violations, the minister writes. ‘Therefore, attention must be paid to human rights in every phase of the life cycle (design, construction, testing, use and monitoring) of AI systems. The AI ​​regulation offers possibilities for this.’ (The EU also recently came up with a proposal for the Date Act

Very special cases

Leijten also asked whether the minister understood that people on benefits automatically feel suspicious when they read that municipalities film the conversations as standard. According to Bruins Slot, the legality of benefits must be checked, but that check must be done with appropriate instruments. The standard filming of conversations will be ‘acceptable only in very special cases’.

Leave a Comment