In an advertising campaign fronted in social media, Sparebank 1 Østlandet writes that “Harald (29) earns money without lifting a finger”, with an accompanying picture of the man. The problem is that Harald does not exist. He is made up by the bank, and his image is generated by artificial intelligence. That the image is AI-generated and that Harald is not real is not disclosed anywhere in the ad. Heri Ramampiaro, head of department and professor at the Department of Computer Technology and Informatics at NTNU, is critical of this use of artificial intelligence. – It is problematic if it is used deliberately to deceive people. – Have stepped a little wrong Siv Stenseth, executive vice president for communications at Sparebank 1 Østlandet, admits that they did not assess well enough before publishing the advertisement. – This is a matter we can only learn from. Here we have stepped a little wrong, says Stenseth to news. She explains that they currently have guidelines not to use content generated by AI externally, and that this is an area they will continue to work on. Siv Stenseth lies flat and says Sparebank 1 Østlandet must learn from the KI blemish. Photo: Anders Bakkerud Larsen / news Now they are going back to old routines. – We will do what we normally do. It is using pictures of real people in our ads. And when we are going to create a customer story, we will use a real customer, says Stenseth. The authorities must step in Here Ramampiaro at NTNU raises the issue and calls for clearer guidelines for the use of AI. Among other things, he believes that what is AI-generated should be labeled. – Things that are generated by artificial intelligence must have a label, and the authorities should step in there. The faster, the better. There is no reason why we in Norway should not be good at this, says Ramampiaro. Researcher Heri Ramampiaro at NTNU makes clear demands to the authorities. AI-generated advertisements must be labeled, he believes. Photo: Press image / NTNU Researcher and author of a book on artificial intelligence, Leonora Bergsjø, thinks the same. – What guidelines and labeling will help us with is to preserve the trust we have in society today. Important actors can help ensure that it is maintained by labeling clearly, so that we are left in no doubt that we can trust what we see and hear. Researcher and author Leonora Bergsjø says labeling of AI-produced material will help us to have trust in large actors, both public and private. Photo: Martin H. W: Zondag / news – Must be clearly stated Stenseth in Sparebank 1 Østlandet supports them in that such content should be labelled. – It must be made clear that it has been generated with the help of artificial intelligence, she says. Ramampiaro believes that when they knowingly and willfully deceive people by presenting it as something else, then it is wrong. Stenseth agrees with that. – We want to appear credible and build trust. And then it must be real customers we use when we create stories, says Stenseth. news spoke to people on the street about the “bank customer” Harald, and they are surprised when they learn that Harald does not exist. See their reactions here! Are people on the street able to recognize Harald, and how do they react when they learn that he does not exist? Digitization minister: Will consider injunction Minister for digitization and administration, Karianne Tung (Ap), believes that artificial intelligence and advertising are a very current issue. She emphasizes that KI is not unregulated. – Content created with the help of AI must comply with the legislation we have today, for example the Personal Data Act, the Act on Equality and the Prohibition of Discrimination and the Marketing Act, she says. These, and all other laws, also apply to the use and development of AI. According to the minister, the government is now working on a new digitization strategy which will, among other things, look at how artificial intelligence will be regulated, developed and used in Norway. Karianne Tung, Minister for Digitization and Administration believes that the use of new technology must be framed. The possibilities in new technology are great, and can make our lives easier and better, create new jobs and solve major challenges, she believes. At the same time, it also brings with it some challenges. Photo: Bernt Sønvisen / Arbeiderpartiet Tung also believes it is problematic if images or other content created by AI pretends to be reality. – In the same way as for illustrative photos or retouched photos, I believe there are factual reasons why this should be marked. That is why I will consider a proposal for a statutory injunction about it, she says. Easy to create Harald Game developer Josef Najjar says it has become very easy to create believable faces via artificial intelligence, like the bank customer Harald in the advertisement in question. – All you need are good programs and a clear order of what you want in terms of appearance. Game developer Josef Najjar in Sarepto studios says it is getting easier to create believable faces via AI. Photo: Knut-Øyvind Hagen / news He believes that technology brings many opportunities for his industry, but that there is also reason to be cautious in the future. – Artificial technology is here to stay, but it is important to try to avoid technology running away from us. Thinks it’s misleading marketing It was Ole Magnus Kinapel, filmmaker at NewsLab and editor for Teknokratiet, who first pointed out the KI ad in a reader post in Aftenposten. Using an AI tool that can recognize other AI-generated images, he found that the image was generated. Kinapel believes that this is misleading marketing from Sparebank 1 Østlandet. – And it is forbidden. If Sonans writes that Fredrik went from two to four in science, it must be true. Because that’s what people trust. Advertising in which the person in the picture does not even exist is experienced as trickery of a new dimension.
ttn-69
Sparebank 1 Østlandet apologizes for the use of advertising generated by artificial intelligence – news Innlandet – Local news, TV and radio
