The European Parliament would like the creators of deepfakes to tag their videos to avoid confusion. In a resolution adopted in January, it calls for labeling of this content. But it seems unlikely that his injunction will be implemented.
It is a very brief passage which appears in the latest resolution of the European Parliament on artificial intelligence, dated January 20, 2021. In the midst of reflections on the guidelines that should be followed on the military and non-military uses of AI, the European parliamentarians briefly addressed the problem of “deepfakes”, these faked and inauthentic videos.
In their resolution, which certainly has no binding legal value, but which at least reflects the concern of the European Parliament on a political level, the elected officials clearly identify the problem that deepfakes can cause – and already cause in part, by force. increasingly realistic realism brought about by advances in computer computing and image processing algorithms.
Thus, MEPs observe that these ultra-realistic special effects ” can be used to blackmail, produce false news reports, damage public confidence or influence public discourse “. They also believe ” what these practices are likely to destabilize states by disseminating false information and influencing elections ».
Deepfake is an English term that has been officially translated into French as “videotox”, a portmanteau word between video and intox. This type of video generally consists of inserting a face in place of another, to give the illusion that another person is being staged. Today, the deepfake is mainly used for pornographic purposes, where celebrities are unwittingly used in porn videos.
Given the current and future drifts caused by this content, elected officials ” express their deep concern »By a lack of supervision. This fear is all the more audible at a time when disinformation, beliefs, alternative facts, information and knowledge rub shoulders, without it being always easy to clearly distinguish them and to recognize their value in relation to each other. to others.
Self-regulation in deepfakes?
But if the stakes are well posed, and the context of a world immersed in post-truth well taken into account, the strategy to respond to them, at least in the short term, is quite puzzling. Indeed, the European Parliament is content to ask ” that the creators [des deepfakes] are required to report When they show videos like this.
As deepfakes consist of ” create forged photos and audio and video recordings “, Which turn out” more and more realistic », And which according to the observation of elected officials are intended for malicious purposes, it seems quite improbable that the creators label this content to warn the public that what he is watching is an extremely thorough editing.
According to the European Parliament, this request particularly targets creators when the deepfake ” pursues electoral goals “. If it does not envisage a ban on this type of technique (would it be respected, anyway?), The Chamber would like its recourse ” is strictly regulated, and that this obligation [d’étiquetage] be rigorously applied ».
On the other hand, much more realistic is the request of the European Parliament for research work ” ensure that technologies designed to curb these phenomena are not left behind by the malicious use of AI “. In short, the shield should not be outclassed by the sword. But in this area, a challenge organized by Facebook showed that the shield is not yet up to standard.
Source: Numerama by www.numerama.com.
*The article has been translated based on the content of Numerama by www.numerama.com. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!
*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.
*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!