Featured Image Credit: X / @immasiddx
Topics: Artificial intelligence, Technology, True Life, Life, Science
Rhianna Benson
Topics: Artificial intelligence, Technology, True Life, Life, Science
Artificial intelligence is at it again.
This week, unnerving advances in image-generation technology by one AI firm have sparked mass fear that photographic evidence will soon be rendered useless in criminal cases.
Nano Banana is an AI image-generation and editing tool launched by Google as part of a subsidiary called Google DeepMind.
Upon its release, the model went viral online for its ability to churn out photorealistic '3D figurine' images.
Advert
Earlier this week, however, a new 'Pro' edition of Nano Banana was made accessible to paying users, allowing them to create visuals so eerily lifelike that tech gurus are sounding the alarm.
New developments allow highly-sophisticated false images to be created, which even the most tech-savvy of AI experts admitted to struggling to detect.
Despite hyper-realistic new 'photographs' enabling faster creative work in areas like advertising, design and education, it is understood to come with a number of disturbing risks.
Nano Banana Pro also allows you to upload an image of yourself, or another real person, and place that individual in virtually any scenario — on a red carpet, in the Antarctic, or even skydiving.
An image generated by X user @immasiddx amassed particular attention for demonstrating the advancements between the original version and Nano Banana Pro.
"It is a step-up in realism specifically," content creator Jeremy Carrasco, who warns his followers of the key things to look out for when assessing potentially AI-generated snaps, told press this week in response to the latest launch.
"So a lot of the things that we used to look for, such as a blurry camera image or something that looked a little bit too glossy or smooth, a lot of that has been straightened out," he told NBC.
Carrasco added that online users must be hyper-alert, especially in 'sensational' cases, or those that are 'strange' and even 'too good to be true'.
What experts are reportedly concerned about, however, is the ability for this advanced technology to place innocent individuals at the scene of a crime - or generate an artificial crime scene altogether.
These developments have also raised concerns over issues like deepfakes, misinformation, identity misuse and trust in digital media.
This is especially said to be the case with regards to high-profile figures.
"I think there should be some expectation that they pull that back," Carrasco explained. "We have to demand they do because just the idea of anyone being able to, essentially, become a PhotoShop Pro overnight and use these celebrities or politicians' likeness is obviously a very frightening thing."
As we say, in light of the release of Nano Banana Pro, a number of online users have spoken out.
"I miss the days when I didn’t have to constantly second-guess whether someone was human or AI," one wrote on Instagram.
Another begged: "Pls God give us Ai regulations before this new version spreads a billion images that we can’t tell are fake anymore."
"I imagine AI will legally need to be disclosed at some point…. Right ?!" questioned a third.
In response to concerns, Google pointed out that its 'Generative AI Prohibited Use Policy' disallows 'generating or distributing content' that, amongst other grievances, 'facilitates 'hatred or hate speech, harassment, bullying, intimidation, abuse, or the insulting of others, violence or the incitement of violence, and 'sexually explicit content, for example, content created for the purpose of pornography or sexual gratification'.
The same policy also outlaws engagement in 'misinformation, misrepresentation, or misleading activities', including, 'frauds, scams, or other deceptive actions, impersonating an individual (living or dead) without explicit disclosure', 'facilitating misleading claims of expertise or capability in sensitive areas - for example in health, finance, government services or the law in order to deceive', 'facilitating misleading claims related to governmental or democratic processes or harmful health practices', and 'misrepresenting the provenance of generated content by claiming it was created solely by a human'.
Google bosses also emphasised that each Nano Banana Pro-generated image includes both an invisible and a visible watermark.