The Minnesota Senate has proposed a bill to ban the use of AI to create fake photos and videos of real people in compromising contexts.
Putin with a resigned look carried away in handcuffs, Donald Trump who tries to oppose the arrest, the Pope with an oversized trapper-style duvet. These are some of the photos that went viral a few months ago, albeit blatantly false. All made with artificial intelligence software, Midjourney.
It is a text-to-image image generator: you simply write what you would like to see and the artificial intelligence transforms the words into images. Just Midjourney had decided to temporarily stop the free trials of its service due to the sudden flood of new users that had led to a system overload.
Other similar programs are Dall-E, Craiyon, Stable Diffusion or Dream by Wombo.
The risks of fake photos and videos
In recent weeks, public opinion has become aware of the potential and risks of artificial intelligence applied to photography and video. In fact, many believed that the widespread images, however unreal, were true. But the problem is not "only" misinformation online.
Because in addition to this kind of images there are also deepfake porn videos, in which the face of a non-consenting person is cropped and used in a video or image with an erotic background, attached for example to the bodies of hardcore actors, generating thus fakes that can create not only embarrassment but also big problems for the victim.
Minnesota: the law against deepfakes
Now the State of Minnesota, in the United States, has decided to criminalize the use of artificial intelligence for the creation of fake news and deepfakes.
The bill aims to hit all those who use artificial intelligence by distorting the image of real people. This applies to both fake news and deepfake porn.
In earlier versions of this bill there were exceptions in cases where the nude image was created for parody purposes only but with the latest version this variant has also been dropped.
The proposal is expected to become law across the country in late May, thus making Minnesota the first US state to adopt a series of laws on this issue.
And Minnesota could become a forerunner for the rest of the world grappling with a heated debate on the responsibility of regulating the use of artificial intelligence.
Original article published on Money.it Italy 2023-05-15 08:00:00. Original title: Gli USA vogliono vietare l’intelligenza artificiale per creare fake news e deepfake
