Meta has entered the AI image-generation race with a product designed to spread across its apps, and with a feature that is already alarming privacy experts. The company has launched Muse Image, an AI tool that creates and edits pictures from text prompts inside Instagram, WhatsApp and its standalone Meta AI app, TechCrunch reported.
What Muse does
Like rival tools from OpenAI and Google, Muse turns a written description into an image, and can edit existing photos. Meta is weaving it through its social apps, where billions of people already share pictures, and plans to let advertisers use it to generate marketing images at scale. For Meta, which has spent heavily to catch up in AI, the appeal is obvious: more ways to create and share content mean more engagement, and more raw material for its advertising machine.
The feature drawing fire
The controversy centers on one capability. Muse lets a user reference other people's Instagram accounts, in effect "@-mentioning" someone, to insert that person's likeness into an AI-generated image, Engadget reported. In other words, you can generate a synthetic picture featuring someone who never posed for it and may not know it was made.
Meta says users remain in control, pointing to existing Instagram privacy settings and "an easy setting to turn this feature off," in its announcement. But that is an opt-out, not an opt-in: the feature works on a person's likeness by default, and the burden of switching it off falls on the potential subject, who must first know it exists. Consent, in other words, is assumed rather than asked for.
Why the timing is fraught
Meta is launching this into a hardening regulatory climate for synthetic media. Data-protection authorities around the world have been warning that AI-generated images of real people, especially nonconsensual or intimate ones, can breach privacy law, and that a person's face is sensitive biometric data that cannot be used freely just because a machine produced the image. The UK's Information Commissioner's Office has cautioned tech firms that AI is no loophole around data-protection rules, as Computing reported.
The concern is not abstract. A tool that can place a named individual into a fabricated scene, deployed on platforms with billions of users, multiplies both the ease and the reach of misuse, from harassment and bullying to defamation and worse, with children and other vulnerable people particularly exposed.
The business calculation
For Meta, the bet is that the benefits, engagement, advertising revenue and staying competitive with OpenAI and Google, outweigh the risks. Muse is offered on a "freemium" model: basic use is free, with heavier use behind a subscription. But the strategy assumes users trust the platform enough to keep sharing, and that assumption sits uneasily with a feature that can depict people without their say-so.
Why it matters
The launch crystallizes a tension running through the whole generative-AI boom: the same capabilities that make these tools compelling, realism, personalization, ease, are the ones that make them dangerous when pointed at real people. For Meta, already facing antitrust and privacy scrutiny, building likeness generation into its core apps with an opt-out rather than opt-in design invites exactly the regulatory attention it can least afford. How Meta handles consent for Muse may end up mattering more to its bottom line than how good the images look. This article is informational and not investment advice.
