Meta has entered the race to build artificial intelligence that makes pictures, launching a tool it calls Muse Image that can generate and edit images from written instructions. The company is aiming the technology at two audiences at once: ordinary users of its apps, and the advertisers whose spending underpins its revenue.
What it does
Muse Image, developed by a Meta AI unit and folded into the company's Meta AI assistant, can turn text prompts into images, restyle photos, add text within pictures and combine several references into one image, Meta said. Basic use is free through Meta AI, while heavier use is reserved for paying subscribers, CNBC reported. The tool is rolling out first in Meta AI, with wider availability across Facebook, Instagram, Messenger and WhatsApp planned in the following weeks, The Next Web reported.
The advertising play
The commercial logic is central. Meta said millions of advertisers already use its generative-AI tools to make ad creative, and Muse Image is meant to slot into that machinery, letting brands spin up on-message variations of an image quickly. For a company whose profits come overwhelmingly from advertising, tools that make ads cheaper and faster to produce, and that keep advertisers inside Meta's ecosystem, are a direct business bet, not just a technology showcase.
A crowded field
Meta is not first here. Rivals including OpenAI, Google, Adobe and specialist tools such as Midjourney already have established image generators and user bases, so Muse Image is a catch-up move as much as a leap. Meta argues its system brings stronger reasoning and better handling of complex requests, though such claims will need independent testing before they can be judged.
The safety questions
As with all AI image tools, the launch raises familiar concerns. Meta said generated images carry invisible watermarks and that people can opt out of having their public profiles used as references. But questions about deepfakes, non-consensual imagery and the copyright status of the data used to train such models hang over the whole industry, and Meta has not spelled out in detail how Muse Image guards against every misuse. Regulators and lawmakers have been pressing platforms on exactly these points. For Meta, the challenge now is not only to catch its rivals technically, but to convince users, advertisers and watchdogs that it can deploy a powerful image machine responsibly.
