Dutch watchdog urges swift implementation of AI rules to protect fundamental rights

By Necva Tastan Sevinc

ISTANBUL (AA) - The Dutch Data Protection Authority (AP) on Thursday urged the government to accelerate the implementation of rules governing artificial intelligence (AI), warning that the rapid spread of the technology poses risks to fundamental rights.

AP Chair Aleid Wolfsen said oversight of AI systems remains insufficient and called on the new government to quickly implement European Union regulations on the technology, the Dutch news broadcaster NOS reported.

“Fundamental rights must be protected,” Wolfsen said, pointing to growing risks including deepfakes, AI-driven fraud, and psychological harm caused by chatbots.

The regulator said the Netherlands should move swiftly to set up a dedicated AI supervisory authority and allocate funding to monitor the use of the technology.

Wolfsen cited the country’s childcare benefits scandal as an example of how algorithms can cause serious harm. Thousands of families were wrongly accused of fraud due to a discriminatory system used by tax authorities.

“Five years after the benefits scandal, the lessons are clear. Anyone who wants to prevent another scandal must act now,” he said.

He also warned about the use of algorithms in probation services to assess whether suspects are likely to reoffend, saying such systems had provided incorrect recommendations.

“Those are decisions about people,” Wolfsen said.

He further noted risks in automated recruitment and selection processes, warning that the use of AI without human oversight could lead to discrimination.

“A decision is made about you, and you don’t know on what basis. Then you can’t go to court to defend yourself,” he added.

The EU has already adopted legislation regulating artificial intelligence, including requirements that developers test powerful AI systems to ensure accuracy and mitigate risks.

However, the Dutch regulator said the rules still need to be implemented at the national level to regulate how AI systems are marketed and used in the Netherlands.

Wolfsen compared the need for AI oversight to safety standards applied to products such as medicines or cars.

“They have to be tested. They have to be safe. Otherwise, it can lead to dangerous situations,” he said.