EU is 'Wild West' compared to US on facial recognition rules

Amazon followed suit a couple of days later putting a temporary, year-long ban on facial recognition contracts with American police departments.

Finally, Microsoft said that they, too, would no longer sell facial recognition to American police departments without federal regulation.

Details aside, these statements all share the implicit confession of the danger that facial recognition poses to human rights and democracy.

This self-containment coming out of Big Tech does not, however, address these very same dangers that exist in the EU.

Although these technologies are used within EU member states as well, the decisions from IBM, Amazon and Microsoft only apply to the American context.

Remarkably, this means that the US now has stricter regulation for facial recognition software than the European Union, although the latter has repeatedly advertised itself as being a role model for AI regulation.

As of today, however, this proclamation has not resulted in any concrete legislation: EU-citizens are not protected from the harmful consequences that facial recognition has been shown to present.

This is not to say that the EU has never dealt with these dangers and how to curtail them, but they have always stopped short of putting anything into effect.

To ban, or not to ban?

For example, at the beginning of this year, the European Commission had drafted their "Structure for the White Paper on artificial intelligence – a European approach", which was leaked to the public.

In this leaked draft, there was a proposed five-year ban on facial recognition technologies.

This temporary moratorium was praised by advocacy groups and watchdogs as a way of giving ample time to craft comprehensive policy governing the use of facial recognition within the EU.

In the official version of the White Paper that was released a month later, however, this ban was nowhere to be found – for reasons unknown to the public.

Instead, the Commission back-pedalled on their suggested regulation and simply classified the use of AI technologies for "remote biometric identification" (e.g. facial recognition) as "high-risk" and wrote that processing of such data can only occur under specific conditions, leaving it up to the discretion of national governmental and law enforcement agencies.

This is by no means enough to control the role that facial recognition software has played and will continue to play in public spaces.

Many contentious topics remain untouched, especially as they pertain to the European context.

For example, what about the use of facial recognition technologies by law enforcement agencies in European states? What about the use of facial recognition in European customs and border patrol?

Or what about the many EU member states who bought access to a database of billions of faceprints collected by a company in the US without citizens' consent or even knowledge, even though it is likely inconsistent with the GDPR?

Currently, Europe is the Wild West; individual states, as well as the Union as a whole, could continue procuring facial recognition software from the private sector and continue using this software in government bodies and law enforcement agencies.

In struggling to come up with a coherent strategy that results in governance and regulation, the EU not only allows tech companies to employ their technologies in Europe – to the detriment of EU-citizens – without being held accountable, it also allows for the private sector to set the tone of the debate around regulation.

PR ploy

Had the EU made good on its many promises to be a leader in AI regulation, weak and ambiguous statements as the ones we heard last week from IBM, Microsoft and Amazon would not be applauded, but rather seen as the public relations ploy they were.

Facial recognition software is not the only instance of AI technologies doing more harm than anticipated, and it certainly won't be the last.

We need to practice better and stronger policy making that holds these actors and these technologies to a standard that accounts for human rights and European values.

In the case of facial recognition, we should reopen the European debate on a moratorium or other concrete regulation as we take a step into that direction.