If you’re going to Germany, be sure you haven’t de-identified your passport photograph

Summary

The German government has banned the use of de-identification technologies on passport photographs to prevent fraudulent use of official documents.

Abstract

In April 2019, the author discussed the effectiveness of de-identification technology developed by Israeli company D-ID, which subtly alters facial features in photos to evade facial recognition systems. This technology, which the author experienced firsthand at the Netexplo Innovation Observatory, modifies facial parameters to protect privacy without altering the image's visual appearance. However, the German government has prohibited such modifications in official documents like passports, fearing that these could allow multiple identities to be linked to a single altered photograph, potentially enabling illegal activities such as unauthorized border crossings. This move reflects the growing tension between privacy protection and the need for secure identification methods.

Opinions

The author views D-ID's de-identification technology as a powerful tool for safeguarding privacy in the era of widespread facial recognition technology.
The German government's ban is seen as a response to the potential misuse of de-identification technology, particularly in the context of immigration and border control.
The author suggests that the ban may indicate a broader trend, questioning whether other countries might follow suit and impose similar restrictions.
There is an underlying concern that the ban could signal a dystopian shift, where measures aimed at enhancing security might infringe on individual rights to privacy and data protection.

If you’re going to Germany, be sure you haven’t de-identified your passport photograph

In April 2019, I wrote an article about Israeli company D-ID, founded by Gil Perry, after meeting him at the Netexplo Innovation Observatory, an event I attend every year.

Perry’s company specializes in de-identification, a process that introduces slight modifications to photographs of faces that cannot be perceived with the naked eye, but which prevent them from being used to feed facial recognition databases. Digital facial recognition is based on a few measurements in an image, so what the D-ID algorithms do is to slightly separate the eyes or modify imperceptibly certain features so they no longer coincide with the original parameters in a photograph, thus preventing algorithmic programs from recognizing them.

As I noted in my piece, this is a very effective way of protecting our privacy: D-ID’s algorithms can easily be applied to any photograph of ourselves we upload to social networks so as to prevent them from identifying us in case they are being used to feed databases, which had happened to me before. Now, the German government has announced a ban on the use of these kind of morphing technologies on photographs for official documents such as identity cards or passports, arguing that these modifications could be used to assign several identities to a single photograph, potentially allowing different people to use the same ID to cross international borders that use electronic verification systems.

From now on, German passport photographs will either have to be taken at official document issuing offices or sent digitally through a secure connection to ensure that they have not been tampered with.

The possibility of using this type of de-identification technology on an official document such as a passport could make sense when visiting certain countries that use facial recognition technology aggressively, although I am unfamiliar with immigration procedures at airports. Obviously, interfering with an official document might be ill-advised, but Germany’s decision to specifically ban the use of this kind of technology seems dystopian, and suggests that its use might indeed be spreading.

Can we expect similar decisions in other countries?

(En español, aquí)