Meta's AI allows third parties to create deepfakes with photos from your Instagram; see how to prevent reproduction Meta announced this week its first AI image generator developed by the company's artificial superintelligence team. Technology, however, allows others to use public photos on Instagram to create deepfakes (see how to stop it). ? Deepfake is a technique that uses artificial intelligence to alter photos and videos. With it, for example, it is possible to swap one person's face for another's or modify what someone says in a video. ?? Do you have any reporting suggestions? Send to g1 Meta's new image generator, called Muse Image, is activated by default and can be used on Instagram, WhatsApp and the Meta AI website, the chatbot from Mark Zuckerberg's company. It will soon come to Facebook and Messenger. g1 tested the tool on the Meta AI website and found that it was capable of creating third-party deepfakes just by mentioning the person's @, without having to inform the social network. In many cases, the AI ??automatically identified that it was Instagram. In others, he even searched for photos of the person on Google.
Meta's AI allows other people to create deepfakes with photos from your Instagram; see how to stop
Meta's AI allows third parties to create deepfakes with photos from your Instagram; see how to prevent reproduction Meta announced this week its first AI image generator developed by the company's artificial...
In its launch statement, Meta said it can use public posts made on its social media channels to generate AI images and said users "have control over how their content can be marked for AI creation, with a simple setting to disable this feature at any time." "Mention a friend with @ in the Meta AI app to bring them into your creation, share directly to your story or group chat, or remix what's trending. It's AI in service of the social experiences that billions of people already love," said the company. g1 also tested the tool with private accounts, but the creation of images was blocked. g1 contacted Meta to find out how the company intends to act in cases of misuse of the tool and is awaiting a response. How to prevent Meta's AI from allowing third parties to create deepfakes with photos from your Instagram; see how to prevent Reproduction/Meta AI Although the feature is being released gradually, Meta already allows you to prevent third parties from using your public content to create images with Meta AI. Here's how: On your profile, tap the menu (the three bars in the top right corner). Under "How others can interact with you," tap "Sharing and reuse." Under "Allow people to reuse your content on Instagram and with Meta's AI features", uncheck the "Posts" and "Reels" options. ?? In the Meta AI app settings, in a separate feature from Muse Image, you can also define who can use your image in AI creations. To do this, access the Meta AI settings, tap "Your image" and send a selfie and other photos of yourself. Then choose who can use these images: "Only you", "Followers I approve", "Followers I also follow", or "Everyone". For Patrícia Peck, a lawyer specializing in digital law, the decision to allow the use of colleagues' images published on Instagram for creations with Meta's AI shows that the race for artificial intelligence is happening without due attention to ethical, governance, security and necessary care for people. "If Meta's AI causes harm to third parties and it is proven that protective measures, so-called guard rails, could have been adopted, but were not due to haste, this could be characterized as recklessness," he stated. WhatsApp gains usernames and will need a number to start a conversation Man sues OpenAI and says ChatGPT reinforced the delusion that he was Jesus Christ
This story was originally published by G1 Tecnologia. Visit the original publication for further details.
Open original publication
