You can stumble across artificial content everywhere on the internet: the Pope in a white Balenciaga jacket, deepfake porn pictures of Taylor Swift, theses written with the voice AI Chat-GPT. They can have dramatic effects: Last May, an AI-generated image of an explosion at the Pentagon on social media caused a brief collapse in the US stock market. This is probably just the beginning: The World Economic Forum, in its annual report on global risks, names AI-generated disinformation as one of the biggest near-term threats. The question therefore arises: Is there now a way to quickly and reliably distinguish AI content from human-generated content?
AI detectors do not deliver what they promise – economy
