Pedophiles are using a popular new artificial intelligence (AI) platform to turn real photos of children into sexual images, it has been revealed.
Parents already know not to upload photos of their children to social media, but many don’t pay much attention. Now things have gotten even worse, with AI allowing some twisted minds to take these photos and edit them in unhealthy ways. So, experts are once again warning parents to be careful about the photos of their children they post online.
Images found on US Midjourney AI Image Generator, which, like ChatGPT, uses prompts to provide a result, although these usually consist of images rather than words.
The platform is used by millions of people and has created such realistic images that people all over the world have been fooled by it, including Twitter users.
For example a photo of him Pope Francis wearing a huge puffy white jacketwas so realistic that it sent social media users into a frenzy earlier this year.
They were too fake footage of Donald Trump’s arrest and ‘The Last Supper Selfie’ created using the platform.
Last year, the platform suffered a backlash when a computer-generated image won first place in an American art competition.
The program recently launched a new version of its software, which increased the photorealism of its imagesonly increases its popularity.
A Times investigation found that some Midjourney users create a large number of sexualized images of children, as well as women and celebrities.
Users go through the Discord communication platform to create prompts, then upload the resulting images to the Midjourney website in a public collection.
Although the company said the content should be “PG-13 and family-friendly”, it also warned that because the technology is new, it “doesn’t always work as expected”.
However, the specific images discovered violate Midjourney and Discord’s Terms of Service.
Although virtual child sexual abuse images are not illegal in the United States, content like this – known as non-photographic images – is banned in England and Wales.
The NSPCC’s Associate Child Online Safety Policy Officer, Richard Collard, said: “It is completely unacceptable for Discord and Midjourney to actively facilitate the creation and hosting of degrading, abusive and sexual depictions of children.
“In some cases this material would be illegal under UK law and by hosting child pornography material they put children at a very real risk of harm.”
He added: “It is incredibly distressing for parents and children to have their images stolen and adapted by criminals.
“By posting photos only to trusted contacts and managing their privacy settings, parents can reduce the risk of images being used in this way.
“But ultimately, tech companies need to take responsibility for looking after how their services are used by offenders.”
In response to the Times’ findings, Midjourney said it would ban users who broke its rules.
Its CEO and founder, David Holz, added: “For the past few months, we have been working on a Scalable AI Supervisor, which we started testing with our user base last week.
A Discord spokesperson told The Times, “Discord has a zero-tolerance policy for the promotion and sharing of non-consensual sexual material, including pornographic material and child sexual abuse material.”
Horror: pedestrians use virtual reality environments
The finding comes amid growing concern about pedophiles exploiting virtual reality environments.
Earlier this year, an NSPCC investigation revealed for the first time how platforms like the metaverse are used to abuse children.
Data showed UK police forces had recorded eight instances where virtual reality (VR) spaces were used for image crimes of child sexual abuse.
Jthe metaverse, which is primarily led by Meta’s Mark Zuckerberg, is a collection of virtual spaces where you can play, work, and communicate with others who aren’t in the same physical space as you.
The Facebook founder has been a key spokesperson for the concept, which is seen as the future of the internet and will blur the lines between physical and digital.
West Midlands Police have recorded five cases of pass abuse and one in Warwickshire, while Surrey Police have recorded two crimes – including one involving Meta’s Oculus headset, now called Quest.
HOW TO SPOT A DEEPFAKE
1. Abnormal eye movement . Eye movements that don’t look natural — or a lack of eye movement, like not blinking — are huge red flags. It’s hard to replicate the act of blinking in a way that feels natural. It is also difficult to reproduce the eye movements of a real person. This is because the eyes generally follow the person being spoken to.
2. Unnatural facial expressions . When something is wrong with a face, it can signal a facial transformation. This happens when one image is stitched onto another.
3. Awkward placement with facial features . If someone’s face is pointing in one direction and their nose is pointing in another, you should be skeptical about the authenticity of the video.
4. Lack of emotion . You can also spot what’s called “face morphing” or image stitching if someone’s face doesn’t seem to show the emotion it should be to match what it is. supposed to say.
5. Strange body or posture . Another sign is if a person’s body shape looks unnatural or if the placement of the head and body is awkward or inconsistent. This can be one of the easiest inconsistencies to spot, as deepfake technology typically focuses on facial features rather than the entire body.
6. Abnormal body movement or body shape . If someone looks distorted or washed out when they turn sideways or move their head, or if their movements are jerky and choppy from frame to frame, you should suspect the video is fake.
7. Unnatural coloring . Abnormal skin tone, discoloration, strange lighting, and misplaced shadows are all signs that what you’re seeing is probably fake.
8. Hair that doesn’t look real . You won’t see frizzy or flyaways. For what; Fake images will not be able to create these individual features.
9. Teeth that don’t look real . The algorithms may not be able to generate individual teeth, so the lack of individual tooth outlines could be a clue.
10. Blur or misalignment. If the edges of the images are blurry or the graphics don’t line up (for example, where a person’s face and neck meet their body), you’ll know something is wrong.
11. Noise or inconsistent sound. Deepfakes typically spend more time on video images than on audio. The result can be poor lip sync, robotic voices, odd pronunciation of words, digital background noise, or even no sound at all.
12. Images that look unnatural when slowed down . If you’re watching a video on a screen larger than your smartphone or have video editing software that can slow down a video, you can zoom in and examine the footage more closely. Zooming in on the lips, for example, will help you see if they’re actually talking or if the lip sync is off.
13. Hashtag discrepancies . There is a cryptographic algorithm that helps video creators show that their videos are authentic. The algorithm is used to insert hashtags at certain points in a video. If the hashtags change, you should suspect video manipulation.
14. Digital Footprints. Blockchain technology can also create a digital fingerprint for videos. While not foolproof, this blockchain-based verification can help establish a video’s authenticity. That’s how it works. When a video is created, the content is entered into a registry that cannot be changed. This technology can help prove the authenticity of a video.
15. Reverse Image Searches . A search of an original image, or a computer-assisted reverse image search, can reveal similar videos online to determine if an image, sound or video has been altered in any way. Although reverse video search technology is not yet available to the public, investing in a tool like this could be worthwhile.