On June 23, 2026, at Cannes Lions, Meta launched an end-to-end creative solution within Ads Manager, centered around a feature called Brand Memory that learns your brand’s identity and tone from your previous ads and generates new creative aligned with that style. In a single workspace, creative and media teams can see which ads are performing, generate new ones based on that learning, test them, and translate them. Meta’s promise is straightforward: your next creative decision should be backed by real performance data, not guesswork.
The headline is the capability; the real story is governance. Because when Meta’s AI generates creative based on your brand, the question stops being “Can AI make my ads?” and becomes “Who decides what gets published, based on what criteria, and under what supervision?” This article separates what is already working from what has only been announced and highlights where you need to keep a human in the loop.
What Did Meta Announce at Cannes Lions 2026 for Meta Ads?
Meta announced an end-to-end creative solution: a unified workspace within Ads Manager where advertising creative can be generated, tested, and translated based on the brand’s own performance history. Nicola Mendelsohn, Meta’s head of global business, presented it on June 23, 2026. The core idea is to close the loop between what is working and what comes next, without jumping between separate tools to generate, measure, and optimize.
Around the main announcement, Meta added several concrete pieces. It expanded AI-powered translation with five new languages for text on images and eleven new languages for video voiceovers, a practical lever for brands entering multilingual markets. It also began testing a Creative Approval Flow integrated into Ads Manager, allowing feedback to happen within the same workflow rather than over email. And it announced the merger of Creator Marketplace and Partnership Ads Hub into a single Meta Creator Marketing Hub, which will arrive later in 2026 and will also incorporate Facebook creators.
Meta backed all of this with a sales figure: it claims that every dollar of advertising investment generates an average return of $4.13, 25% higher than in 2022. This should be attributed carefully: it is a figure reported by Meta based on a study commissioned by Meta itself (April 2026, involving more than one million campaigns), not an independent audit, and the increase coincides with the rollout of Advantage+, making the cause difficult to isolate. Use it as context, not as a proven fact.
What Is Brand Memory and How Does It Work?
Brand Memory is the feature that allows Meta’s AI to learn your brand’s identity, tone, and creative patterns from your existing ad library and apply that learning to the creative it generates. Instead of starting from a blank page every time, the system starts with something it already knows about your brand: which messages worked, what visual style you use, and how your communications sound.
In practice, this addresses the problem that made AI-generated creative feel generic. A generator that does not know your brand produces ads that could belong to anyone; Brand Memory aims to produce ads that resemble yours. For small teams without a full creative department, this is a genuinely useful change: it multiplies production while maintaining a foundation of brand consistency.
The limitation matters just as much as the promise. Brand Memory learns from what you have already published, so it inherits both your successes and your bad habits: if your historical library is weak or inconsistent, the system scales that weakness. The quality of what the AI generates is constrained by the quality of what you feed it. At JP Director, as an AI-powered audiovisual production company, we treat the brand library as an asset to be curated, not an archive to be accumulated: what goes into Brand Memory determines what comes out.
Can I Use Brand Memory on My Account Today?
Not generally yet. As of June 23, 2026, Brand Memory is being tested with a limited group of selected advertisers, and the broader rollout is expected “in the coming months,” according to Meta, with no firm date. Several of the headline features from the announcement are still in testing or labeled “coming months,” so before building a strategy around them, you should verify their actual availability and the status of your account in Ads Manager.
This is where an issue that almost no coverage emphasizes appears: governance. Meta’s AI creative features tend to activate by default through opt-out enrollment, meaning they can be enabled without you explicitly turning them on, leaving you responsible for disabling them if you do not want them. That reverses the burden: instead of consciously deciding to use AI, you have to discover that it is already being used.
The practical consequence is simple and urgent: check your account’s enrollment status before assuming anything. A feature that changes your creative without your knowledge is a brand risk, not a convenience. The first step is not “start generating with AI”; it is to audit what is currently enabled on your account and deliberately decide what you want to leave on.
Why Does This Change the Role of Creative in Meta Ads?
It changes the role because, in Meta in 2026, creative is no longer merely the wrapper around your targeting: it is the targeting signal. With the Andromeda engine, Meta reads the elements of your ad—visual, message, and format—to decide who should see it, rather than having you manually select audiences. Brand Memory adds the missing layer: now AI not only helps decide who sees the ad, but also helps generate it. If you have already been following how AI is increasingly managing Facebook Ads campaigns, Brand Memory is the next link in that same chain.
The result is a shift in role: from executor to strategist and auditor. When the machine builds the ad, chooses the audience, and adjusts the bid, human value is no longer about pressing the right buttons. It lies in the inputs AI cannot generate on its own: the quality of conversion signals, creative direction, brand boundaries, and strategic interpretation of performance across platforms.
This does not eliminate the value of the marketer or agency; it concentrates it. What becomes automated is production and distribution; what remains scarce is judgment: knowing which concept is worth scaling, which message protects the brand, and when AI’s default behavior conflicts with business objectives. That is precisely the boundary where an agency stops “managing budget” and starts directing the machine.
What Should AI NOT Do on Its Own?
AI should not publish creative without human review to ensure that the ad remains truthful, compliant, and safe for the brand—especially when content is being produced faster than a creative team could previously review it. The speed of generation is precisely the risk: when dozens of variations can be produced in minutes, the bottleneck is no longer production; it is approving them with sound judgment. In regulated industries such as healthcare, finance, or legal services, an AI-generated variation can violate compliance rules without anyone noticing until it is already in circulation.
That is why creative direction matters more, not less, in the Brand Memory era. Generating ten on-brand ads is easy; knowing which of those ten tells the right story within the first two seconds, connects with the right stage of the customer journey, and does not cannibalize your own messaging—that remains human work. The principles behind creating ads that actually convert do not change because AI generates the draft; they change hands, from producer to director.
To illustrate the risk with a hypothetical case—not a real audit, but a typical scenario—imagine an account that enables AI generation without curating its library or establishing an approval workflow. Brand Memory learns from old ads containing an outdated claim, replicates it across fifteen new variations, and those variations enter the auction before anyone reviews them. The cost is not merely creative; it is a brand and, in a regulated industry, compliance cost. A human approval layer would have stopped all fifteen at one point in the process.
What Should You Do With Meta Ads Now?
Before allowing Meta’s AI to generate your creative, do three things in this order: audit what is enabled on your account, curate your brand library, and establish a human approval workflow. The order is not accidental: without auditing enrollment, you may be using AI without knowing it; without curating the library, Brand Memory can amplify your worst ads; and without an approval workflow, the machine’s speed becomes a risk rather than an advantage.
Specifically: first, check the enrollment status of AI creative features in Ads Manager and deliberately decide which ones you want enabled. Second, select your strongest historical ads—the ones that represent your brand at its best—because they are the raw material Brand Memory will learn from. Third, establish a human review point before publishing, especially if you operate in a regulated industry. Fourth, if you sell to multilingual audiences, take advantage of Meta’s expanded AI translation capabilities to adapt creative to Spanish-, English-, or other-language markets without rebuilding every asset from scratch.
This is where an agency with both creative and technical expertise provides real value. At JP Director, we combine AI-powered audiovisual production with Meta Ads management: we curate the brand library that feeds Brand Memory, build the approval workflow that protects compliance, and keep creative direction in human hands while the machine scales production. The goal is not to use more AI; it is to use it with judgment, so that every generated ad remains unmistakably yours.
Frequently Asked Questions
Is Meta’s Brand Memory Already Available to All Advertisers?
No. At the time of the Cannes Lions announcement on June 23, 2026, Brand Memory was being tested with selected advertisers, and Meta indicated a broader rollout “in the coming months” without a confirmed date. Before planning a strategy around this feature, verify its actual availability and your account’s status directly in Ads Manager, because several of the announced capabilities are still in testing.
Can Meta’s AI Start Changing My Ads Without My Authorization?
It is possible, which is why you should check. Meta’s AI creative features tend to use opt-out enrollment, meaning they can be enabled by default and leave you responsible for disabling them. The practical recommendation is to audit the enrollment status of AI features on your account and consciously decide which ones you want enabled, rather than assuming everything is disabled until you turn it on.
If Meta Generates Creative with AI, Do I Still Need a Creative Team or Agency?
Yes, but their role changes. AI automates production and distribution, but it does not independently generate the inputs that determine the outcome: the quality of conversion signals, creative direction, brand boundaries, and the judgment required to decide which concept to scale. The human role shifts from executing campaigns to directing and auditing the machine, which is where a creative-focused Meta Ads agency concentrates its value in 2026.
Is It True That Meta Ads Returns $4.13 for Every Dollar Spent?
That is a figure reported by Meta, not an independently audited fact. It comes from a study commissioned by Meta in April 2026 involving more than one million campaigns, and the reported increase since 2022 coincides with the rollout of its Advantage+ suite, making it difficult to isolate the cause. It is useful as context for the direction of the market, but it should be cited as a claim made by Meta rather than as a guaranteed return for your account.
