Meta’s AI-Creative Default and the Control Problem
Meta’s latest wave of AI ad creative tools promises smarter performance, but some advertisers say the reality feels more like losing the steering wheel. Clothing brand Snag Tights noticed what it called a strange “AI sheen” on its Meta ad creative, only to discover that AI tools were modifying images without the brand’s knowledge. Its CEO argued that when visuals no longer depict real products, customers can feel misled, especially if AI imagery suggests items that never existed. Agencies have also reported unexpected changes to image size, backgrounds, and layouts, alongside budgets being funneled into AI ad testing they didn’t explicitly plan. Meta says advertisers can opt out of AI creative testing in Ad Account Settings and that tests only touch a small share of impressions. But marketers say identifying, diagnosing, and consistently disabling these AI interventions can be complex in day-to-day operations.
Unwanted Treatments, Broken Tone, and Brand Risk at Scale
The core frustration with AI ad creative on large platforms is not that AI exists, but that it behaves unpredictably. Marketers describe visual distortions, odd crops, or altered backgrounds that subtly break carefully crafted art direction. Even when images remain on-brand, text or layout tweaks can clash with a brand’s tone of voice, especially for regulated or safety-sensitive categories. At scale, this becomes a governance problem: global campaigns depend on strict brand guidelines, consistent typography, and compliant disclosures. When AI experiments vary formats or messaging on its own, reviewers struggle to keep pace. In practice, this can mean discovering off-brand or misleading ads only after they’ve run. For brands invested in long-term equity, AI in advertising that behaves like a black box—especially when switched on by default—feels less like optimization and more like an uncontrolled risk surface.
Canva AI 2.0 and Starti AI Studio: Automation with Guardrails
While some marketers fight to turn off auto-generated Meta AI ads, other platforms are positioning AI as a controllable partner. Canva AI 2.0 recasts the design tool as an “agentic platform” that stays with teams from idea to execution. Its conversational design and layered object intelligence allow users to prompt AI, then refine every element—copy, layout, and assets—without losing editability. A memory library learns preferences, reinforcing brand consistency over time. Starti AI Studio 2.0 takes a similar systems-first approach for video, framing its Video Agent as a creative collaborator that handles shot planning and structure while keeping outputs fully adjustable. Its motion graphics engine generates reusable, editable components and automatically preserves brand visuals across scenes. Both tools emphasize workflow continuity and post-campaign analysis, pitching AI ad creative as an iterative, transparent system rather than opaque auto-optimization—precisely the kind of control marketers say is missing from some default marketing automation tools.
Why Platforms Default to More Automation—and Where It Collides with Brands
Platforms lean into aggressive automation for clear reasons: more AI ad creative variants mean higher ad volume, richer A/B testing, and potentially better performance. Systems like Meta’s Advantage+ combine creative tweaks with AI-driven targeting and budget optimization to chase incremental lift over millions of impressions. For platforms, the business logic is compelling: automation can lower friction for smaller advertisers and consolidate complex decisions into a few toggles. But this model can collide with brands that must enforce legal, regulatory, or brand safety constraints. Unapproved treatments may introduce non-compliant claims, missing disclaimers, or culturally sensitive imagery. Even when nothing goes wrong, automated experimentation can erode the consistency that long-term brand building depends on. The underlying tension is strategic: performance marketing automation optimizes for short-term metrics, while brand and governance teams optimize for predictability, accountability, and human oversight over AI in advertising.
Practical Steps: Balancing AI Performance with Brand Control
Advertisers looking to harness AI ad creative without losing control can start with a few pragmatic moves. First, audit platform settings: identify any default AI creative testing, dynamic formats, or Advantage-style bundles, and document which toggles control them. Ask platforms explicit questions: How are AI variants labeled in reports? Can specific campaigns or accounts be fully exempt from AI-generated assets? What review or approval steps exist before AI images go live? Second, build internal workflows that treat AI as a draft engine, not an auto-pilot. Use tools like Canva AI 2.0 or Starti AI Studio where teams can iteratively refine outputs, lock brand elements, and connect creative to performance analytics. Finally, define clear no-go zones for AI—such as regulated claims or sensitive product lines—while allowing automation in lower-risk campaigns. The goal is not to reject marketing automation tools outright, but to combine human creative direction with AI that stays visibly, reliably under brand control.