DEEPFAKES: HOW REAL IS REALITY?
Author: Wilfried Platten, editor and IT expert at the Munich PR and communications agency PR-COM, which specialises in high-tech
Fake news is not an exclusive phenomenon of modern times. Disinformation, propaganda and myth-making are ancient. However, with deepfakes, i.e. AI-generated texts, images and videos, we are now facing a completely new, more perfidious form of misinformation and disinformation. It therefore has much greater potential to become a threat to democracy. The election campaign in the USA has given us an unappetising and indigestible foretaste of what awaits us with videos faked using GenAI. So the crucial question is: how real is the reality that is being presented to us in the media around the clock? Not so long ago, the press took over the answer to this question: As the fourth estate, it was functionally important in its responsibility for the curated dissemination of information, sorting out lies and deception in a reasonably independent and reliable manner. However, it has lost this function, replaced by social media, where every narrative, no matter how absurd, finds a platform and followers.
GenAI exacerbates this problem of a missing corrective. Because it is getting better and better, the ability to expose fakes is increasingly being lost. GenAI has now reached the point where the deep fake dazzle generated with its help can only be unmasked by itself. A technology is thus fundamentally beyond human control. And what happens to the DeepFakes afterwards? They don’t disappear into oblivion. As long as they cannot be recognised and filtered out, they will remain part of the big data that is needed on such a massive scale for the development of AI applications. Deep fakes are thus gradually and subliminally integrated into the billions of texts, images and videos generated by GenAI. They are then no longer identifiable as such, spread like a creeping poison and become part of our reality – with incalculable consequences for liberal societies.