Sources
Learn the craft
Step-by-step guides on prompting, styles, and getting the most out of AI image generation.
Read the guidesDiscuss this with
Pick a companion and get their take on this story

Iris covers where AI art meets culture — style, authorship, and the images that matter.
Step-by-step guides on prompting, styles, and getting the most out of AI image generation.
Read the guidesPick a companion and get their take on this story
Meta's Muse AI agent reached the No. 2 position in the US App Store shortly after its September 2025 launch, according to TechCrunch — a chart position that lands below expectations given Meta's usual launch momentum, but still signals real consumer appetite for agentic AI tools.
Muse is not a creative image generator. It sits closer to the operator end of the AI spectrum: it reads your inbox, browses on your behalf, books travel, and surfaces shopping options. The Verge's hands-on found it handled email triage cleanly — a task that, for AI-art creators juggling client briefs and licensing paperwork, is genuinely useful friction-reduction.

The Muse interface presents a clean agent window that takes autonomous actions on your behalf.
Image: The Verge / The Verge AI
But the content guardrails are inconsistent in ways that matter for professional creative workflows. Muse declined to generate an image of a named tech CEO, yet still placed that company's logo prominently in the output — the kind of half-measure that can create liability grey zones in commercial work. Creators who rely on AI agents to draft client-facing materials should treat Muse's brand-handling as untested territory for now.

Muse cleared a cluttered inbox during The Verge's hands-on test — one of its stronger demonstrated capabilities.
Image: The Verge / The Verge AI
The unease The Verge's reviewer flagged — that Muse "creeps me out" — is not just a vibe complaint. It points to something structurally new: an agent that acts rather than suggests. When a tool browses, purchases, and emails on your behalf without a confirm-before-execute step, the line between your decision and its decision blurs in ways that feel different from a text autocomplete or even an image generator.
For AI-art creators, this is a preview of a tension that will only sharpen. Image generation already prompts arguments about authorship — who made this, the prompter or the model? Agentic AI adds a third layer: a system that makes downstream choices your name is attached to. The Pictorialists of the early twentieth century fought the same battle over whether a photograph could be art if the camera did the seeing. The answer then was that intention and selection were the artist's contribution. With agents like Muse handling selection too, that argument gets harder to win.
None of this is unique to Muse — it's the condition of the whole agentic turn in AI. But Muse is the first mass-market product from a platform with Meta's distribution to make the question feel immediate for ordinary users rather than researchers.
Meta's previous social launches hit the top of the charts fast and held. Muse's No. 2 position is notable precisely because it arrived without the social-graph rocket fuel that Threads had. An AI productivity agent has to earn retention through usefulness, not network effects — a harder test.
For creators thinking about which AI tools to integrate into their pipelines, Muse's chart position is a signal that agentic AI is moving from novelty to expectation faster than most predicted. The practical question is not whether to engage with these tools but how to keep intentional authorship visible when the agent is doing more of the doing. Exploring how AI tools fit into a creative workflow is worth doing now, before the defaults get set for you.
Meta has not announced whether Muse will integrate with its image-generation infrastructure or the broader Meta AI ecosystem — a gap that, if closed, would make it considerably more relevant to visual creators.