EU-US roadmap on AI risks interfering with EU’s artificial intelligence regulation

BEUC and the Transatlantic Consumer Dialogue (TACD) – a network of 77 consumer and digital rights organisations – are concerned about the EU-US ‘joint roadmap’ on artificial intelligence (AI) developed as part of the Trade and Technology Council. Ahead of a public event focused on the roadmap today (16/02), we are calling on policymakers to make sure the roadmap does not interfere with how AI is regulated in either the EU or US.

While AI products and services can improve consumers’ lives, the deployment of AI systems can have major negative implications for people’s privacy, their capacity to take autonomous decisions, and their ability to hold business responsible if something goes wrong.

BEUC and TACD point out a number of issues with the proposed EU-US AI roadmap including:

• The roadmap foresees that the EU and the US will jointly define AI concepts such as ‘harm’ and ‘bias’. However, these concepts are not yet defined in EU legislation and are currently under discussion by lawmakers. If these concepts are defined between the EU and the US in the Trade and Technology Council, there is a risk that the roadmap pre-empts domestic legislative processes.

• Most consumer groups in the EU and the US have not been involved in the stakeholder consultation during the drafting of the roadmap. This raises concerns as the roadmap’s activities will impact consumers and therefore require broader stakeholder engagement.  

Monique Goyens, BEUC Director General and TACD co-chair, commented:

“This proposed EU-US roadmap goes too far in trying to define key artificial intelligence concepts. In its present form the roadmap could interfere with how we regulate artificial intelligence, a process which should be left to domestic legislators who are democratically accountable.

“Consumer and digital rights organisations call upon EU and US policymakers to instead focus their cooperation on how regulators and civil society can access algorithms and source codes in order to detect discriminatory, unfair, biased, and other illegal conduct. We also urge policymakers leading the Trade and Technology Council to better involve civil society in the future.”

Read more

• BEUC’s work on artificial intelligence
• TACD recommendations on the EU-US cooperation agenda