A moratorium is not going to help us win the battle

Recently we’ve seen numerous organisations calling for a moratorium on face surveillance. The most heard arguments include the lack of demonstrable necessity, the biased and inaccurate nature of the technology, and finally the absence of a regulatory framework. However valid these arguments are, we believe they aren’t going to help us win this battle.

Necessity

First, the focus on “necessity”. We wholeheartedly agree with the European Data Protection SupervisorRead the blog of the EDPS about facial recognition that any interference in our fundamental rights must be demonstrably necessary, and convenience and efficiency do not amount to a demonstrable need. However, how many surveillance measures have we seen being introduced with a poorly motivated, let alone demonstrated, necessity? Unfortunately, we believe we desperately need more in terms of protection.

Technological deficiencies

Secondly, there’s the question of inaccuracy and bias. Our worry as regards to arguing for a moratorium on the basis of this concern, is that the technological deficiencies might be solvable over time, at least to an extent that brings the percentage of false positive and negatives within the realms of what our political leaders deem acceptable. More importantly, however, is that we might just be looking at a technology that becomes more dangerous, the better it works. Not when it’s giving you access to your phone, but definitely when applied as a mass surveillance tool.

Regulatory framework

Finally, calling for a regulatory framework, might imply to some that current legislation is ambiguous about the acceptability of face surveillance. We need to be very clear that assessing face surveillance in the light of the European Convention on Human Rights, the Charter of Fundamental Rights of the European Union, and the principles set out in the General Data Protection Regulation, do not leave space for the deployment of facial recognition surveillance in public space, since it requires mass-scale processing of biometrics. Period.

Why we need to be bolder

More than anything else, what a moratorium will result in, is time. Time in which the technology will become normalized. Time in which industry will deploy its lobbyists. Time in which the companies at the forefront of product development, search for and find product-market-fit. Time in which civil society will again and again mobilize citizens until those citizens become fatigued and weary, and disbelieving that their voice makes a difference.

We’re concerned, therefore, that the demand for a moratorium isn’t bold enough. We believe existing regulation needs to be enforced, banning the deployment of facial recognition as a surveillance tool in public space.

The costs to our liberties are too big

First, there’s no question that we’re talking about a mass surveillance tool that will severely limit our rights and freedoms. No balancing act, or proportionality assessment, can make the infringement caused by face surveillance fair or just.

It is incompatible with our data protection framework

Second, facial recognition surveillance in public space, in our view, is clearly incompatible with our data protection framework. In other words: face surveillance is illegal. Biometric data are extremely sensitive and due diligence requires limiting the processing of this data as much as possible. As we know, facial recognition surveillance in public space inherently requires mass-scale processing of biometric data. A study from 2016 shows that half of all United States adults are already included in a law enforcement facial recognition database. In the Netherlands it is already 1 in every 12 adultsLees hier wat we daarover zeggen bij RTL Nieuws (Dutch). We cannot opt-out or avoid public space. We cannot change our face, or leave it at home. So let’s treat our face as something valuable.

Never underestimate a good function creep

Finally, we are concerned we will not be able to contain the use of face surveillance. History has taught us never to underestimate a good function creep. There are several ways the use and effects of facial recognition surveillance might expand over time. First, the legal basis and/or the scope of the basis can be expanded. Limiting the use of such far-reaching technology to combating terrorism might sound limited, but the limitation and therefore protections are dependent on government classifications. Several examples around the world, including in the Netherlands, show that even non-violent citizen interests groups are classified as ‘extremist’ or ‘terrorist’ when more powers to surveil these groups are desired. A second example of how function creep will take place, is with regards to access to the data. Waiving the fraud-prevention-flag, and showing a complete distrust of citizens, government institutions are very keen to share access and combine databases. Why would facial recognition databases be exempt from this data hunger?

Stop (offline) tracking

Calling for a moratorium is risky. We believe we need to be bolder. We are in agreement, for the most part, that our human rights and data protection frameworks protect people against tracking online. We now need to agree that they do the same offline.