The Dutch College for Human Rights warns against the use of algorithms for LinkedIn and job board recruitment. According to the human rights organization, algorithms can lead to discrimination. The College urges employers and governments to pay more attention to the practice.

The College for Human Rights’ research indicates that almost all companies use algorithms to recruit staff. LinkedIn or other job boards provide algorithms to distribute vacancies and actively search and approach candidates. Some companies use their own algorithms for recruiting and selecting candidates.

Dangerous development

The College for Human Rights finds the development risky. According to its researchers, many companies don’t know how these algorithms work, which can lead to discrimination and exclusion.

For example, when posting a vacancy on LinkedIn or other job sites, employers don’t know who sees the vacancy and who shows up in a search query. Algorithms can assess applicants faster and on a larger scale. This can lead to bias, according to the human rights organization.

Companies that use proprietary algorithms in employee recruitment and selection processes are oblivious to the risks as well. In more than half of the cases, the study indicates, the companies don’t know whether automatic selection processes are reviewed and whether certain individuals are unfairly rejected. Nor do they know why applicants are being rejected, as they hardly ever test the various recruitment technologies for bias.

Measures to be taken

The College believes that companies should review why algorithms cause specific candidates to be left out and, if so, whether this is justified. The Board also urges companies to inform applicants of the usage of algorithms in advance.

Furthermore, the College argues that the Dutch government should draw up additional regulations to ensure that everyone has a fair chance at jobs and to combat labor market discrimination. The College calls on the government to make it easier for employers to test algorithms for bias. Suitable means are lacking at this time.

Tip: LinkedIn’s new tools help visual content creators share their work