JAKARTA - Meta released Muse Image, an artificial intelligence model that allows others to create highly realistic images with the faces of Instagram users. The risk mainly lurks for users over 18 years old who use public accounts.
CNET quoted Saturday, July 11, reported that anyone with a Meta AI account can enter an Instagram username into the image creation command. The result can convincingly display the user's face, even though the image is not a photo of the actual event.
In this case, deepfake refers to an AI-generated image that mimics a person's face very realistically. The person whose face is used may not even realize that he or she has been included in someone else's AI creation.
Muse Image is the first creative model from the Muse Spark family developed by Meta's superintelligence lab. The model is designed to handle more complex commands, combine multiple photos, and edit existing images.
The service is already available on the Meta AI, Instagram, and WhatsApp apps. Meta also plans to bring it to Facebook, Messenger, and advertiser services.
CNET tested the feature by entering another Instagram username. A deepfake image can be created in less than a minute.
The ease immediately triggered criticism. The actor's union SAG-AFTRA told the LA Times that the decision to automatically include users in the feature was a big mistake in reading the public's attitude towards the dangers and losses that could arise.
Executive Director of the National Center on Sexual Exploitation Haley McNamara also questioned how the feature could have been approved.
"How can this be given the green light? Meta creates a clear and foreseeable opportunity for exploitation, sexual violence, harassment, and identity fraud," he said, quoted by CNet.
The risk is very real for creators and influencers. Their names and faces are not only personal identities, but also part of brands and sources of income.
Meta states that Muse Image has a safety system to prevent the creation of illegal, abusive, or defamatory content. However, CNET notes that safety systems on other AI models have previously been bypassed by malicious users.
How to Limit Face Use on Instagram
Users can choose two ways. The first step is to change the Instagram account to private.
1. Open Instagram and log in to the profile page. 2. Tap the three-line icon in the upper right corner. 3. Select Account Privacy. 4. Activate Private Account.
The content of a private account cannot be accessed by anyone to be processed or used in the creation of AI content. However, accounts are also more difficult to find and interactions with other users are more limited.
Creators, influencers, or business actors who still need a public account can turn off the permission to reuse content.
1. Open Instagram and log in to the profile page. 2. Tap the three-line icon in the upper right corner. 3. Scroll down to the Sharing and reuse section. 4. Turn off the option to Allow people to reuse your content on Instagram and with AI features on Meta.
The setting limits the use of uploads and Reels by Meta's AI feature. However, this step cannot be considered a guarantee that all forms of abuse will disappear.
Meta AI app users can also open Settings, then select Your likeness. In that section, users can specify who can use their likeness, from just themselves, approved followers, users who follow each other, to everyone.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
