AI advertising is no longer optional, but compulsory – at least as far as Meta is concerned. At the Cannes Lions 2026, the tech giant announced that with its new AI tools, "the threshold has been crossed": in the future, ads are set to generate themselves, test themselves and optimise themselves, directly within Ads Manager. For businesses in Austria, this means: anyone who still thinks AI advertising on Meta is just a nice extra for the major players has another think coming.

What did Meta actually announce in Cannes?

On 23 June 2026, Meta unveiled an end-to-end AI advertising ecosystem at Cannes Lions – and pulled out all the stops right away: a study covering more than 1 million campaigns reveals that every dollar invested in Meta advertising returns an average of $4.13 in revenue, up 25 per cent since 2022.

At the core of the announcement is an end-to-end creative AI solution featuring "Brand Memory": the AI learns how your brand sounds and looks from your previous ads, and subsequently produces new variants independently, tests them against each other and automatically rolls out the winners. Added to this is the new Meta Creator Marketing Hub, which merges the Creator Marketplace and Partnership Ads Hub – over 5 million Instagram creators can already be found there, with Facebook creators set to join. Moreover, with the Meta Business Agent Platform, businesses can build their own AI assistants that sell directly in WhatsApp and Messenger: according to Meta, more than 1 million businesses are already active there.

In plain English: at Meta, AI is no longer an add-on tool within Ads Manager – it is Ads Manager.

AI no longer builds your ads on demand. It tests, learns and decides – whether you have a say or not.

Is AI advertising really better – or just cheaper?

The $4.13 figure sounds impressive, but it comes from a study funded by Meta itself. A degree of caution is warranted: this is a provider assessing its own product. Meta CMO Alex Schultz commented on LinkedIn that more than 8 million advertisers are now using the platform’s generative AI tools and that even experienced creatives could no longer reliably tell AI-generated ads apart from human-made ones in direct tests.

Just how quickly things can go wrong was demonstrated during that very same week: outdoor retailer REI was enrolled by Meta into an AI image editing tool without consultation – the result was a distorted bicycle ad that ran unnoticed on Instagram for roughly a week. The new "Brand Memory" feature is Meta's answer to precisely this governance challenge. Nevertheless, the bold premise holds: AI advertising is becoming the standard, whether you like having control over it or not.

Is AI advertising as worthwhile for small businesses as it is for corporations?

Yes, and for a simple reason: AI advertising lowers the barrier to entry most significantly for anyone who hasn't previously had their own graphic design or copywriting team. Instead of manually producing ten ad creatives, a CI briefing is all it takes – the AI uses it to generate dozens of variants for different target audiences and languages. For large brands, this means faster international rollouts; for local businesses, it means a small advertising budget finds the creatives that actually work much faster through automated testing.

In Austria, the foundation for this has long been established: according to a recent survey by KMU Forschung Austria commissioned by the WKÖ, 44 per cent of domestic companies are already actively using AI, with another 31 per cent planning or testing it. The biggest hurdles are data protection concerns (46 per cent), unclear legal frameworks (39 per cent) and a lack of specialist expertise (34 per cent) – not a lack of willingness. This aligns with what we at oha! are currently seeing among clients: the demand for AI-driven ads is there, but so is the uncertainty regarding implementation.

The key takeaways in 30 seconds

→

Meta introduced an end-to-end AI advertising ecosystem at Cannes Lions 2026, declaring AI to have transitioned from an experiment to the standard.

→

At its core: an end-to-end creative AI solution featuring "Brand Memory", which learns from existing ads and automatically generates and tests new variants.

→

The new Meta Creator Marketing Hub combines creator discovery and ad activation, with more than 5 million Instagram creators already listed.

→

In Austria, 44 % of businesses are already actively using AI; the main bottlenecks are legal uncertainty and skill shortages, not a lack of willingness.

→

Brand guardrails are essential before launching a campaign: the REI incident demonstrates how quickly uncontrolled AI advertising can backfire.

How do you get started in practice?

Five steps that we at oha! are currently implementing with clients:

1. Assess the starting point. Analyse how heavily you already rely on Meta Ads today and where your greatest leverage lies: reach, leads or sales.

2. Define a pilot project. Select a single campaign, such as customer acquisition, and test automated creative generation there in a targeted way.

3. Establish brand guardrails. Define your logo, colours, tone of voice and no-gos before the AI begins advertising on your behalf – the REI story shows why this is not merely a nice-to-have.

4. Test creator collaborations systematically. Use the new Creator Marketing Hub as soon as it becomes available to identify 1–2 suitable creators and test their content as ads.

5. Document results. Track whether AI variants genuinely beat your click-through rate or conversion rate by at least 10 per cent before scaling up its use.

AI in advertising is no longer a competitive advantage that you can simply opt into like an optional add-on. It has become part of the rules of the game that Meta is currently rewriting – much like ChatGPT, Gemini, Claude and others have ensured that copywriting is no longer left solely to creative departments. Those who choose guardrails over flying blind right now hold the stronger cards.

Frequently asked questions about AI advertising on Meta

01

What does "Crossing the AI Threshold" mean for Meta?

Meta uses this to describe the point at which AI-generated ad creatives perform at least as well on average as manually produced ones – at lower costs and higher speed. At Cannes Lions 2026, Meta declared that this point has now been reached.

02

Do I have to use AI ads now?

No, these are currently tools provided within Ads Manager as options, not obligations. However, Meta is clearly positioning AI as the new standard workflow – those who ignore it forfeit speed and cost advantages over their competitors.

03

Are AI-generated ads as good as those created by humans?

According to Meta CMO Alex Schultz, even seasoned creatives could no longer reliably distinguish between AI-generated and human-made ads in head-to-head testing. However, the quality relies heavily on the underlying brand guardrails.

04

What is the Meta Creator Marketing Hub?

A new, centralised interface that brings together Creator Marketplace and Partnership Ads Hub. Companies can use it to find creators, deploy their content directly as ads and analyse performance. The launch is planned for later in 2026.

05

How secure are my brand guidelines with automatically generated advertising?

Not automatically secure. The case of outdoor retailer REI – which was enrolled into an AI image editing tool without consent, leaving a distorted ad running unnoticed for around a week – demonstrates that default settings need to be actively vetted. This is precisely where the new "Brand Memory" feature is designed to step in.

06

How much does it cost to get started with AI-driven Meta Ads?

The AI features are included in the existing Ads Manager. Additional costs arise primarily from testing budget and time invested in setting up brand guardrails, rather than from a separate licence fee.