Artificial Intelligence is one of the hottest topics when it comes to recent technological developments. How can it be tamed, how can it be used – and, does it have the power to eventually replace us?
Not only in the European Union but also on a global scale, the subject has been high on the agenda for political leaders, public and private sector figures and the general public, warning about its potential hazards and the need to regulate. The journalistic profession has also been following the subject closely.
Photoshop and CGI fakes in Hollywood films have been around for decades. With artificial intelligence, it is much easier, cheaper and faster to create false claims – it is democratising the ability to fake.
This image of Pope Francis in a Balenciaga puffer coat has been widely circulated on the web. Participants learned what to look for in AI-generated images.
Source: PopBase/X
Stefan Voß, head of verification at the German Press Agency dpa and board member of the European Fact-Checking Standards Network (EFCSN) and Patrick Neumann, head of dpa-academy and recruitment officer, facilitated the full-day workshop.
Around 25 participants from 13 news agencies participating in the European Newsroom project learned about how current programmes can be used to create or altertexts, images, audio and videos and learned that AI-generated images are images – and not (fake) photos.
In practical examples, participants learned about the inner and outer logics which could unmask an artificially created image. Clues to look for could include: Are hands, ears and hair looking off? Are there any inconsistencies in collars or jewelry? Can other recordings of the scene be found? Does the reaction of onlookers seem plausible or rather strange?
Throughout the workshop, the facilitators and participants lively discussed the potentials, pitfalls and dangers of artificial intelligence in journalism.