"STOP Meta from secretly testing weird AI features on your ads with your money," Peter Quadrel wrote on X this week. His post, verified through xurl, had 135 likes, 186 bookmarks, and 12,909 impressions when we checked it. The complaint was not about AI creative in theory. It was about waking up to find Meta testing an AI voiceover on live ads.
That is the kind of seller pain we pay attention to. DTC teams are not rejecting automation. They are asking for control. When a brand has spent months tuning its hook, product claim, visual style, and offer, a small creative change can turn a working ad into something that feels off-brand.
What changed
Meta's own docs confirm the shift. Advantage+ creative enhancements can adjust the media and text an advertiser uploads. Some enhancements may be turned on by default, and Meta tells advertisers to check ad previews and review which enhancements are active before publishing.
The list is getting broader. Meta labels several Advantage+ creative enhancements as AI features, including image animation, AI-generated music, and overlays. Meta also says some enhancement settings may only appear inside Advanced preview, not the main Ad creative section.
That detail matters. If the control is hidden one layer deeper than the place an operator normally reviews ads, QA becomes a process problem rather than a media buying preference alone.
The operator lens
For a DTC brand running Meta ads, the question is no longer "should we use AI creative?" The better question is: which parts of the creative system can change automatically, and who checks them before spend goes live?
We would audit this in five places:
- Campaign objective and format: different formats expose different enhancement options.
- Enhancement status: what is on by default, what is manually selected, and what only appears in Advanced preview.
- Brand fit: whether generated music, overlays, animation, or text variations still match the product and offer.
- Landing page match: whether the ad promise still lines up with the page, price framing, and product detail page.
- Post-publish drift: whether edits or Meta updates turn features back on after the first QA pass.
The risk is subtle. A generated overlay may increase attention but soften the claim. AI music may make a static ad feel more native in Reels but clash with a premium product. A text variation may technically preserve the "core message" while changing the buying reason.
None of these are automatically bad. They are also not automatically safe.
Why this matters for smaller teams
Large brands can build review layers around every creative change. Smaller DTC teams usually cannot. The founder, media buyer, designer, and retention person may all be the same two people on a busy week.
That is why default-on creative optimization is a workflow issue. Meta is moving toward more automated ad assembly because the system has more inventory, placements, and creative surfaces than a human team can manually tailor. The platform wants more variants. Operators want variants that do not break the brand.
A practical rule: treat every AI creative enhancement like a junior editor. It can help create versions faster. It can also misunderstand tone, exaggerate the product, or make the ad feel cheaper than the brand. Review it like you would review a first draft.
How we think about this at Auxora
Auxora is built for this exact gap between platform automation and operator judgment. We do not think every founder should become a Meta Ads settings expert. Brands also should not hand the wheel to default settings and hope the algorithm understands their positioning.
Our view is simple: AI should do more of the checking, while a human-standard operating model decides what passes.
For a Meta account, that means a weekly creative QA sweep that looks for changed enhancement settings, mismatched ad-to-page promises, off-brand generated assets, and weak creative variants before they eat budget. It also means documenting what was allowed, what was turned off, and why.
That documentation compounds. After a few cycles, the brand has a clear rulebook: use AI music only for certain placements, block overlays on premium product shots, allow image animation on lifestyle assets, require manual review for generated text. The system becomes faster because judgment is captured instead of repeated from scratch.
What to check today
If you are running Meta ads this week, open one active campaign and check the Advantage+ creative section. Then open Advanced preview. Look for anything labeled AI, anything turned on by default, and any creative change that would surprise your team if a customer saw it first.
If the answer is "I am not sure," that is the signal. Your ad account does not need less AI. It needs a cleaner QA loop.
Want a second set of eyes on your Meta and Google funnel? Run a free Auxora GTM report and we will check where automation is helping, where it is drifting, and what should be reviewed before your next spend cycle.
Sources
- X post by Peter Quadrel, May 27, 2026, verified with xurl metrics.
- Meta business docs: "About Advantage+ Creative."
- Meta business docs: "Turn Off Advantage+ Creative Enhancements in Meta Ads Manager."
- Meta for Business: "Cannes Lions 2025: Introducing The Next Era of Generative AI for Advertisers and Agencies."
