After the opening of artificial intelligence tools to the general public, we have seen a lot of more or less positive phenomena emerge on the web. Among those that are to be considered as deviations of the modern world, we can talk about deepfake. Many celebrities have been victims of it and staged in disturbing situations. Today, it is the candidates of the American elections who are threatened by this scourge. While they will be held on November 5, 2024, hyper-tampering is multiplying, to the point of fear that it will influence the final results.
What is a deepfake?
Translated as “Deep fake” in French, deepfake is a photo or video montage that aims to make a character say or commit what he has never done. With AI, the possibilities are very vast and these illusory fabrications can be disconcertingly realistic. The creators of these chimeras are able to implant images and voices that clearly correspond to the imitated person, leaving a wide range of resources to spread various fake news. Among the most popular tools, we can mention Sora from OpenAI or Stable Diffusion from Stability.
As one might expect, these chimeras are therefore misused by the various members of political parties to make the opposing party say words that it has never said, thus putting it in an extreme embarrassment in front of the general public. Fortunately, as we will see, it is quite possible to recognize them if one consults the publisher’s guide toExpressVPN and that we have a minimum of critical thinking.
What deepfakes are to be feared in the US elections?
Deepfakes are numerous and are already actively circulating on the net at present. Among them, we can count:
-
Fake accounts created on social networks: Many fake accounts have already been spotted and they are likely to gradually increase as the D-day approaches.
-
Photomontages depicting embarrassing acts: We can find photos unfairly distributed where we see candidates or members of their parties posing in incongruous places, with personalities considered “unsavoury” by the population or in positions that could tarnish their reputation.
-
Video montages in which a candidate or his supporter can address thorny subjects: These are the most feared. They are often those that are considered to be the most credible. However, they are created from scratch using powerful algorithms capable of searching for different information across the web so that the result corresponds to reality. Thus, we can see fairly realistic videos, with a voice, gestures and face totally similar to the real entity.
The danger of these practices is that voters can not only be fooled by this misleading content, but they can also start to develop a distrust of everything they see and completely lose confidence in everything they are shown. This will inevitably lead to a higher abstention rate than in previous years.
How to quickly spot a deepfake?
Fortunately, there are a lot of methods to spot a deepfake more or less quickly. First of all, you need to sharpen your critical thinking. If a news story seems too alarming or unbelievable to you, it could be the work of AI. Before spreading it, check its veracity. To do this, try to find the source, see if it is well reported in official media and try to trace the account that issued it and its history. If you notice a questionable practice, do not take it into consideration. In addition, many tools designed by large technology companies also have the ability to spot certain flaws in deepfake content and inform you about them. Finally, if you come across fake news that is too obvious, do not hesitate to report it. The State has set up certain provisions to combat this phenomenon and could well punish the perpetrator.