The Netherlands must do more to fight deepfakes, expert argue

The Netherlands needs to make more of an effort in fighting deep fakes, VVD MEP Bart Groothuis told BNR. According to numerous experts and politicians, the country is lagging behind in detecting manipulated media in which the face and/or voice of a person has been altered using artificial intelligence software to give the appearance of authenticity.

Primarily, Groothuis argued social media companies need to become legally obliged to warn their users against deepfakes. “This way political disinformation cannot go viral“, he argued. 

Stefan Leijnen, professor of artificial intelligence at the Hogeschool Utrecht, stated that no software will ever be fully capable to detect deepfakes. That is why other measures are also needed to prevent abuse, such as proper legislation and informing the general public on the issue.

Experts previously stated the software meant to bust deepfakes first needs to be vastly improved. Tom Dobber, who is researching deep fakes at the University of Amsterdam, stressed the same. According to him, such software is still in its early phase and needs to be improved drastically before it can used to combat deep fakes properly.

“This software is hardly able to distinguish good and bad deepfakes from real videos. There's still so much to be done there,“ he said.

Stricter rules to prevent the abuse of deepfakes are currently being considered on the European level. Labour MP Barbara Kathmann still argued that the Netherlands should take the initiative to explicitly prohibit abuse by amending Dutch law. As an example, she cited California which previously banned political deepfakes during election season.

According to Dobber, the national legislation may not serve effectively in case of the deepfakes, due to their creators often being residents of other countries rather than the Netherlands. Similar to other experts, he advocated awareness campaigns that teach people what deepfakes are and what consequences they can have.