This week’s edition is Cannes Lions 2026. Less a creative festival this year, more a public auction of where the next decade of performance marketing actually gets built. OpenAI used the week to confirm it is openly in the ad business. Google, Amazon, TikTok and Meta all announced platform shifts that move AI from a creative gimmick into the buying layer. And Reddit quietly positioned itself as something more interesting than a social platform. Six platform stories follow, all from one week on the Croisette. The brands that get clear about which of these shifts matter to them will move faster than the ones trying to chase all of them at once.
OpenAI used Cannes to confirm it is now openly in the ad business
OpenAI used Cannes Lions 2026 to pitch ChatGPT advertising directly to brand marketers, with the move widely read as a signal ahead of its expected IPO. David Dugan outlined more of the playbook at the festival, including chat-native ad formats and the expansion of a UK ads manager with CPC pricing, bringing the channel into territory that performance marketers already understand. Adweek reported that fewer users are dismissing ChatGPT ads as the format scales, which is an early but meaningful signal that the integration is feeling natural rather than disruptive. StackAdapt has also dropped its minimum spend commitment for ChatGPT ads, removing one of the last practical barriers for mid-market advertisers. Taken together, this is no longer a beta channel for global brands with experimental budgets. It is a live auction that is about to get competitive.
What this means: ChatGPT advertising has moved from concept to live channel inside six months, and the window to learn it cheaply is closing. The UK ads manager launch with CPC pricing means the mechanics are familiar, but the auction dynamics are not yet understood, which is exactly when early movers benefit most. The signal worth taking seriously is users dismissing ads less often as the format scales; that is the opposite of what usually happens when a new ad format is introduced at volume, and it suggests the placement is earning attention rather than interrupting it. For performance marketers, this is a new auction to learn before it gets expensive. The competence gap between brands that test now and brands waiting for case studies will open quickly and will be hard to close.
Reddit positions itself as the trust layer between AI recommendations and the buy button
At Cannes this week, Reddit unveiled two new ad formats: a free-form ad and a bespoke image creation ad, both designed to feel native to the platform’s community-led environment. Alongside the formats, Reddit released a data point that deserves more attention than it has received: nearly half of shoppers say they verify AI-generated product recommendations on Reddit before making a purchase. That figure positions Reddit not just as a social platform but as a structural part of the AI-mediated purchase journey. It is the place consumers go to sense-check what an algorithm has told them. The new formats give brands a way to be present in those verification moments rather than absent from them. For considered-purchase categories, this changes the strategic case for Reddit investment significantly.
What this means: this is the most interesting platform story of the week, even if it did not dominate the Cannes headlines. As AI-generated recommendations become more common, through ChatGPT, Gemini, Perplexity and AI overviews in search, consumers are looking for human verification before they commit to a purchase. Reddit, with its anonymous, opinionated, search-discoverable conversations, is structurally well-suited to that role in a way that most other platforms are not. For brands in considered-purchase categories (tech, beauty, finance, automotive), being visible and well-represented inside Reddit threads is no longer optional. The recommendation may come from AI. The decision is increasingly being made on Reddit.
Google is rolling Gemini deeper into YouTube and reframing how brands buy attention
Google’s presence at Cannes was built around Gemini, with new Gemini-powered YouTube tools that let brands create, adapt and target video creative inside the platform rather than upstream of it. The announcement sat alongside Google’s $75m DeepMind investment in A24, framed as a bet on AI-assisted film and series production, and a defensive World Cup search campaign that Digiday described as a “brand counterattack” against shifting consumer search behaviour. Performance Marketing World reported the YouTube changes as the most significant platform-level shift Google has made this year, with creative variation, audience matching and measurement now sitting inside a single Gemini-assisted workflow. The direction of travel is clear: Google is collapsing the gap between creative production, media buying and measurement into one AI-assisted environment. The question for brands is whether their teams and partners can move at that speed.
What this means: Google’s Cannes story is not the headlines. It is the workflow. By embedding Gemini directly into YouTube creative and buying, Google is removing the friction between producing a video, testing variations and putting paid behind the winning version. For performance marketers, this changes how YouTube campaigns get built. Briefs that used to assume two-week creative cycles can now flex week to week, and the brands that adapt their planning rhythm to match will get more out of the same budget. The A24 deal is the more strategic signal. Google is betting that high-quality AI-assisted content is going to be a defensible advantage, not a commoditised one, and that has implications for every brand thinking about where its own creative bar needs to sit.
Amazon announced this week that it is pushing conversational AI ads beyond its own properties and onto the open internet, with new formats that let advertisers run AI-driven product conversations across publisher inventory rather than only inside Amazon’s owned environments. The Retail Ad Service also launched in the UK this week, giving British retailers and brands access to the same ad infrastructure Amazon has been refining in the US for the last two years. Alongside both moves, Alexa is being repositioned as a shopping agent that can act for advertisers outside Amazon, signalling that Amazon’s ambition is to be the buying interface for AI-mediated commerce wherever it happens. Adweek and Digiday both flagged the open-web push as the most consequential platform development of the week. The implication is that Amazon is no longer just a retail media network. It is positioning itself as the conversational commerce layer of the open internet.
What this means: Amazon’s three Cannes announcements add up to a coherent strategy that is bigger than any one of them. Conversational AI ads on the open internet, Retail Ad Service in the UK, and Alexa as a multi-retailer shopping agent are three different routes to the same goal: making Amazon the default interface between an AI-mediated query and a purchase. For UK brands, the Retail Ad Service launch is the most immediate practical change. The same ad infrastructure that Amazon brands in the US have been using for two years is now available, and the brands that get familiar with it first will set the benchmarks for what good looks like. The longer-term competitive question is whether Amazon’s open-web conversational ad layer becomes the new default for AI commerce, in which case being absent is more expensive than being early.
TikTok has launched a new AI ad agent and is building out a creator partnerships playbook
TikTok brought a new AI ad agent to Cannes Lions 2026 that automates campaign setup, creative variation and optimisation inside TikTok Ads Manager, alongside three new partnerships designed to scale brand and creator collaborations through structured networks rather than one-off briefs. MediaPost and Performance Marketing World both reported the announcement as TikTok’s most direct move yet to compete with Meta’s automated ad infrastructure on functionality and cost. The platform also unveiled custom creator networks built around vertical categories, letting brands access curated creator pools matched to specific audience segments and campaign objectives. The combined effect is a TikTok ad stack that is closer to a self-serve performance platform than a social-creative one. For brands selling to younger audiences, the question is no longer whether to be on TikTok. It is whether your existing buying rhythm can keep up with what the platform is now capable of doing automatically.
What this means: TikTok’s AI ad agent is the platform’s most significant performance-side announcement of the year. Automated creative variation, automated campaign setup and automated optimisation all sitting inside one platform tool means the lift required to run TikTok at scale just dropped. For mid-market brands that have treated TikTok as a creative experiment line, this is the moment it becomes a real performance channel. The custom creator networks are the more strategic shift. They make creator activation feel less like a one-off project and more like an audience targeting capability, which changes how briefs should be written. Brands still treating creator partnerships as a separate, one-deal-at-a-time activity are now buying creator content at a structural disadvantage to those using the new infrastructure.
Meta launched a complete AI creative ad ecosystem and is betting $145bn on it working
Meta launched what it is calling a complete AI creative ad ecosystem this week, with end-to-end tools spanning idea generation, video and image production, audience-matched creative variation, and direct deployment into Facebook and Instagram inventory. Performance Marketing World framed it as Meta’s most ambitious creative product launch in years, designed to justify the $145bn in AI capital expenditure Meta is on track to spend this year. The platform also rolled out new Live Video Ads and an AI Mode for Facebook that pulls from public information across its platforms to power smarter discovery and targeting. Adweek reported the moves as Meta betting that owning the full creative-to-conversion stack will defend its ad business against the encroachment of OpenAI, Google and Amazon. For advertisers, the practical question is whether the new creative tools genuinely lower the cost of producing high-performing variations, or whether the savings get absorbed into platform fees and inflation.
What this means: Meta is making the largest bet on AI creative production of any platform this year, and the strategy is now visible. Owning the full pipeline from idea to deployed creative inside Facebook and Instagram lets Meta capture more of the value that currently sits with separate creative tools and production partners. For brands, that is potentially good news on cost and speed. The risk is the same one that always comes with platform-native creative: the work tends to look like the platform rather than the brand. The discipline that matters now is treating Meta’s AI creative tools as a way to test more, faster, while keeping the brand layer firmly outside the platform. The brands that get that balance right will see real efficiency gains. The brands that hand the whole creative process to Meta will get cheaper outputs and weaker brand assets.
That’s a wrap for this week. Potential Unpacked drops weekly, helping you keep up with what’s changing and what to do about it.
Want to chat? Drop us a line at hello@26pmx.com.