Meta launched Muse Image on July 7, 2026 — the first AI image model to come out of Meta Superintelligence Labs, the division led by Alexandr Wang. The rollout across consumer apps (Instagram, WhatsApp, and Meta AI) immediately drew backlash from artists and privacy advocates over training data and consent. Within days, Meta pulled the standalone consumer image-generation feature.
The Advantage+ integration was not part of that pullback. According to Meta's official announcement on about.fb.com, Muse Image is set to power advertiser-specific image generation inside Advantage+ creative "in the coming weeks" — a timeline that started July 7 and extends through Q3 2026. The ads version ships regardless.
What the model actually does in Advantage+
Muse Image replaces Meta's previous AI creative layer in Advantage+, which had documented quality problems and produced assets that frequently looked off-brand. The replacement is more capable in three specific ways, based on Meta's announcement and reporting from ContentGrip and Venti Scale.
It generates contextual backgrounds around your product images. Instead of showing your product on a plain white background, the model creates lifestyle scenes and category-appropriate environments around your existing product photos. A skincare product gets a bathroom vanity. A piece of furniture gets a living room.
It builds lifestyle variations from your current ad creative. If you're running a hero shot of a product, Muse Image can generate versions showing it in use, styled differently, or framed for different audience segments — without a photoshoot.
It produces static images from your video creative. Brands running video-led Advantage+ campaigns can have the model extract and refine still assets from existing footage for use in static placements.
All three outputs run through Advantage+ Creative, which is default-on for most Advantage+ Shopping campaigns. That detail matters. This isn't a new opt-in tool you select in campaign setup. It's being integrated into the infrastructure that was already running your ads.
Why the previous AI layer is not the comparison
Meta's AI creative tools in Advantage+ have been around for a while — mostly as background expansion utilities that produced inconsistent results. The industry read on them was fairly negative: generated assets looked generic, underperformed against strong human creative, and occasionally produced things that didn't represent the brand.
Muse Image is a different tier of model. Reporting from Venti Scale and ContentGrip describes it parsing creative briefs with more specificity, adjusting individual visual elements rather than regenerating wholesale, and producing variation sets with fewer revision cycles. "Your June tests are already stale," Venti Scale noted — meaning the creative performance data you gathered before Muse Image started rolling into accounts may not predict what happens after.
What that means in practice: the performance signal from Advantage+ creative testing is in flux. Results from experiments run against the old model won't transfer cleanly to the new one.
The visibility gap most brands are sitting in
Meta reports that more than 8 million advertisers are already using its AI creative tools, according to figures cited by ContentGrip. Most of them aren't actively managing which enhancement layers are running at any given moment.
Advantage+ creative reporting shows performance by asset, but there's no clear label that distinguishes AI-generated variations from the creatives you uploaded. If Muse Image starts generating and serving new product image variations, those assets appear in your reporting feed. You see the performance number. You don't necessarily see that the image wasn't one of yours.
This isn't a hypothetical problem. It's the architecture Meta is building. The consumer-facing version of Muse Image got pulled because users didn't want an AI generating their personal photos without an explicit opt-in. The Advantage+ version doesn't have that opt-in conversation. It's rolling into your account in Q3, and the input you've given the system determines what it produces.
What to do before the rollout hits your account
Two things worth doing now rather than after.
First, get a snapshot of your current Advantage+ creative performance. Export asset-level reporting from any active Advantage+ Shopping campaigns. You want to know which images are serving, at what frequency, and against which audiences. That's your before state.
Second, review your creative brief language in Advantage+. If your brand and product description is vague — "modern lifestyle" or "clean aesthetic" — the model has wide latitude to interpret what that means visually. Specific constraints work better: product context, what's in or out of frame, what colors the brand doesn't use. The more specific your inputs, the less room the model has to improvise in directions you wouldn't choose yourself.
If you want a clear picture of what your Advantage+ campaigns are actually doing right now, before the creative stack underneath them changes, the free audit at Gromerce runs in a few minutes.
The consumer feature got pulled because people didn't want Meta's AI generating their personal images. The Advantage+ integration doesn't have that conversation. What you've told the system about your brand is what it'll build from.
Sources: Meta for Business, The Verge, TechCrunch, MediaPost, July 2026
