Editor’s Note: This story is part of ‘Systems Error’, a series by CNN’s As Equals, investigating how your gender shapes your life online. For information about how CNN As Equals is funded and more, check out our FAQs.
A European human rights body has ruled that Facebook’s algorithm shows gender bias when promoting job advertisements, marking what activists say is a crucial step in holding big tech companies accountable for the design of their platforms.
The Netherlands Institute for Human Rights said in a February 18 decision that Facebook’s algorithm reinforced gender stereotypes by mainly showing “typically female professions” to female Facebook users in the Netherlands and that Meta (META), the social platform’s owner, should have monitored and adjusted its algorithm to prevent that.
The Institute’s decision follows CNN As Equals reporting revealing that Facebook users in Europe were missing out on job opportunities due to gender bias.
The 2023 article was based on findings shared with CNN by international non-profit Global Witness, which investigated Facebook’s job ads and found that ads in the Netherlands and five other countries often targeted users based on historical gender stereotypes.
For example, ads for mechanic positions were predominantly shown to men, while those for preschool teacher roles were primarily directed to women. Global Witness said its experiments in the Netherlands, France, India, Ireland, the United Kingdom and South Africa demonstrated that the algorithm perpetuated similar biases around the world. The non-profit’s investigation led to four complaints from the Dutch human rights group Bureau Clara Wichmann and the French organization Fondation des Femmes.
The Netherlands Institute for Human Rights said in its February ruling that Meta Platforms Ireland Ltd., which manages Facebook ads in Europe, failed to demonstrate that its advertising algorithm does not engage in prohibited gender discrimination. The Dutch body said Facebook must revise its advertising algorithm to prevent further discrimination.
The European Union has several directives that prohibit discrimination based on gender, including in online advertising.
The Institute’s ruling added that “Meta Ireland has acknowledged that the gender data point can be part of the algorithm. Meta Ireland has not refuted that this data point can promote stereotyping via the algorithm.”
A Meta spokesperson told CNN that it would not be commenting on the matter.
Meta spokesperson Ashley Settle previously told CNN that the company applies “targeting restrictions to advertisers when setting up campaigns for employment, as well as housing and credit ads.” Those audience targeting restrictions are in place in the United States, Canada and more than 40 European countries and territories, including France and the Netherlands, according to Meta.
“We do not allow advertisers to target these ads based on gender,” Settle said in a 2023 statement. “We continue to work with stakeholders and experts across academia, human rights groups and other disciplines on the best ways to study and address algorithmic fairness.” Meta did not respond to questions from CNN at the time about how the algorithm that runs its ad system is trained. In a 2020 blog post about its ad delivery system, Facebook said ads are shown to users based on a variety of factors, including “behavior on and off” the platform.
Berty Bannor of Bureau Clara Wichmann celebrated the Dutch institute’s decision, telling CNN that the ruling was significant.
“Today is a great day for Dutch Facebook users, who have an accessible mechanism to hold multinational tech companies such as Meta accountable and ensure the rights they enjoy offline are upheld in the digital space,” Bannor said.
“I see this as a first step in showing that anti-discrimination laws apply just as much to big tech companies as they do to the offline world,” she added.
Rosie Sharpe, Senior Campaigner on Digital Threats at Global Witness, said the ruling “marks an important step towards holding Big Tech accountable for how they design their services and the discriminatory impact their algorithms can have on people.”
“We hope this ruling can be used as a springboard for further action, in Europe and beyond,” she added.
While the decision by the Netherlands Institute for Human Rights is not legally binding, experts say that, if the case is further escalated, a court will be required to consider the Institute’s findings.
Dutch lawyer Anton Ekker, who specializes in artificial intelligence and digital rights, told CNN that the Institute’s ruling could lead to fines by the Dutch data protection regulator or orders to modify specific algorithms, specifically those that reinforce inequalities and disproportionately harm marginalized groups based on gender, race, ethnicity or religion.
If Meta does not take action on the job ads algorithm, NGOs might choose to pursue further legal action to stop the discriminatory use of its algorithms, he said.
The Dutch ruling comes as protection of digital rights has been severely undermined, particularly for women and marginalized groups, Bannor said.
Last month, Meta said it would end its diversity, equity and inclusion programs, change its policies on hateful conduct on its platforms and drop its third-party fact-checking programs in the US.
Users are now allowed to, for example, refer to “women as household objects or property” or “transgender or non-binary people as ‘it,’” according to a section of the policy prohibiting such speech that was removed. A new section of the policy notes Meta will allow “allegations of mental illness or abnormality when based on gender or sexual orientation, given political and religious discourse about transgenderism and homosexuality.”
Previously, such comments would have been subject to removal by Meta’s moderators.
Meta has faced various allegations of discrimination over the past decade, including lawsuits in the US regarding housing, employment and credit ads. As a result, the company has modified its algorithm for these ads in the US.
Sharpe at Global Witness told CNN that it’s “outrageous” the same changes were not applied globally, arguing that algorithms and AI are increasingly impacting everyday life and posing significant risks to social justice.
CNN’s Kara Fox contributed reporting.
