Conclusions

• Author: European Union Agency for Fundamental Rights (EU body or agency)
• Pages: 33-34

FRA Focus

33

Conclusions

Using facial recognition technology – a technology

that has been developing quickly in the past years

and is increasingly used by multiple actors – affects

a range of fundamental rights. However, there is

limited information about the way and extent to

which the technology is used by law enforcement,

and about the impact of its use on fundamental

rights. Working with new AI-driven technologies,

which are not yet fully understood and where not

much experience has yet been gathered, requires

the involvement of all relevant stakeholders and

experts from different disciplines.

Facial images constitute biometric data, EU law rec-

ognises, as they can be used to identify individuals.

Facial recognition technology can be used in many

different ways, such as verifying the identity of a

person, checking whether a person is among a list

of people, and even to categorise people accord-

ing to different characteristics. Live facial recogni-

tion technology detects all faces on video footage

and then compares the faces against watch lists –

potentially used at public spaces.

While not much information is available about the

actual use of facial recognition technology in the

EU, several Member States are considering, test-

ing or planning the use of the technology for law

enforcement purposes. Most actively, the police in

the United Kingdom carried out several tests in real

life situations such as sports events, even using real

watch lists. Other law enforcement agencies tested

the accuracy of the technology in larger tests with

volunteers, such as the police in Berlin, Germany

or in Nice, France. The lack of more comprehensive

information about the actual use of the technology

limits the opportunities to analyse its fundamental

rights implications. In particular, there are no laws

or other guidance or information on who will be

included in potential watch lists.

The fundamental rights implications of using facial

recognition technology vary considerably depending

on the purpose, context and scope of the use. Some

of the fundamental rights implications stem from the

technology’s lack of accuracy. Accuracy has strongly

increased, but the technology still always comes

with a certain rate of error, which can negatively

impact fundamental rights. Moreover, importantly,

several fundamental rights concerns would remain

even if there were a complete absence of errors.

Notwithstanding the varying context, purpose and

scope of the use of facial recognition technology,

several fundamental rights considerations apply. The

way facial images are obtained and used – poten-

tially without consent or opportunities to opt out

– can have a negative impact on people’s dignity.

Relatedly, the rights to respect for private life and

protection of personal data are at the core of fun-

damental rights concerns when using facial recog-

nition technology. In addition, any use of the tech-

nology needs to be thoroughly assessed in terms of

its potential impact on non-discrimination and rights

of special groups, such as children, older persons

and persons with disabilities, because of the (some-

times unknown) varying accuracy of the technology

for these groups and according to other protected

characteristics. Moreover, freedom of expression,

association and assembly must not be undermined

by the use of the technology.

Lastly, the paper highlights that it is essential to

consider procedural rights when facial recognition

technology is used by public administrations, includ-

ing the right to good administration and the right

to an effective remedy and fair trial.

Given the novelty of the lechnology as well as the

lack of experience and detailed studies on the impact

of facial recognition technologies, multiple aspects

are key to consider before deploying such a sys-

tem in real life applications:

Following the example of the large-scale EU IT

systems, a clear and suff‌iciently detailed legal

framework must regulate the deployment and

use of facial recognition technologies. Determin-

ing when the processing of facial images is nec-

essary and proportionate will depend on the pur-

pose for which the technology is used and on the

safeguards in place to protect individuals whose

facial images are subjected to automated pro-

cessing from possible negative consequences.

Forms of facial recognition that involve a very

high degree of intrusion into fundamental rights,

compromising the inviolable essential core of

one or more fundamental rights, are unlawful.

A distinction must be made between the pro-

cessing of facial images for verif‌ication purposes,

when two facial images are compared to verify

if they appertain to the same person; and their

processing for identif‌ication purposes, when a

facial image is run against a database or watchlist

of facial images. The risk of interferences with

fundamental rights is higher in the second case

and therefore the necessity and proportionality

test must be stricter.



So-called “live facial recognition technologies”−

when facial images are extracted from video

cameras deployed in public spaces – are particu-

larly challenging. Such a use triggers different