Welcome to Week In Review
From new tools transforming creative work, to campaigns reshaped by automation, we track the biggest shifts at the intersection of AI, media, and marketing. This week, publisher ad supply falls up to 40 percent in Q2, OpenAI’s ad revenue forecast looks shaky under analyst scrutiny, and Publicis reports its AI-powered services now make up 87 percent of revenue.
Top Stories of the Week
Publisher Ad Supply Fell by Up to 40% as AI Search Impacts the Open Web
New benchmarking data from Ozone, reported by Digiday, found that publisher ad request volumes were down by roughly 32 percent to 37 percent year-on-year in the US and 39 percent to 41 percent in the UK between April and June 2026, tracked across approximately 20 billion impressions. Despite that steep decline in display ad requests, spend hasn’t collapsed at the same rate, with rising eCPMs – particularly in the UK, where they were around 30 percent higher year on year in June – doing most of the work to plug the gap left by disappearing volume.
Why it matters:
The 40 percent supply drop is the clearest hard evidence yet that AI search is not just changing how people discover content – it is actively dismantling the economic model that has underpinned open web publishing for fifteen years. Every AI Overview that answers a query without a click is an ad request that never gets made. The fact that eCPMs are rising as supply falls offers publishers a partial cushion, but it also highlights that there is a limit to how far price can compensate for volume before buyers simply redirect budgets to walled gardens where measurement is cleaner and inventory more abundant.
Stagwell Launches Its Own AI-Powered Media Curation Marketplace
Stagwell is set to launch Stagwell Curate, an AI platform that pulls curated CTV, online video, display and audio ad inventory from publishers and adtech partners into its own central reservoir, in an attempt to reduce the company’s reliance on “ubiquitous” marketplace solutions like The Trade Desk’s OpenPath. The system uses AI agents built in Claude and data pulled from The Trade Desk’s OpenSincera for a “domain health check” to create bespoke packages suited to a given client’s brief via a process the agency owns rather than a third-party “black box.”
Why it matters:
Curation has historically been the adtech industry’s territory — a layer of intelligence that agencies paid partners to provide and largely took on trust. Stagwell bringing it in-house, powered by AI agents and independent domain quality signals, is a direct challenge to that arrangement. It is also a bet that the competitive advantage of a media agency in an agentic era lies not in access to inventory — which is increasingly commoditised — but in the quality of the decisioning layer that sits above it. If AI can do the curation work faster, more transparently and more cheaply than a third-party platform, the case for paying adtech partners a tech tax to do it weakens considerably.
OpenAI’s Ad Business Is On Pace to Miss Its Own Forecast by 90%
Adweek reported that OpenAI has projected $2.5 billion in ad revenue this year and $100 billion by 2030, but Emarketer’s data finds that standalone chatbots like ChatGPT, Microsoft Copilot, Google AI Mode and Amazon Alexa for Shopping will generate less than $1 billion in ad revenue this year, and just $5.41 billion by 2030. The forecast assumes OpenAI captures search ad budgets en masse from traditional search sellers, dominates a fully mature chatbot ad market, and outperforms every ad format in history – all at once.
Why it matters:
OpenAI’s ad business is only six months old, and the gap between its own projections and what independent analysts think is achievable is enormous. That doesn’t mean ChatGPT ads won’t matter – it means the current narrative around them is running well ahead of the evidence. For marketers, the sensible position is to test and learn now rather than commit significant budget to a platform that has yet to prove it can deliver at scale. The hype is real; the results are still to come.
Quote of the Week
Brands & Agencies
Tefal Taps Creators and AI to Fuel Product Innovation
The kitchen appliances company Tefal has launched a programme inviting creator communities to co-develop product ideas and campaigns using AI tools, positioning its creator partnerships not just as a distribution channel but as a mechanism for consumer insight and R&D, Marketing Week reported. The approach reflects a growing trend among FMCG brands experimenting with AI-assisted co-creation as a route to faster, more market-responsive product development.
Publicis Reports 87% of Revenue Now Comes From AI-Powered Marketing
Publicis Groupe reported in its first-half 2026 results that its AI-powered marketing services, representing 87 percent of total net revenue, delivered 6.5 percent organic growth in Q2, driven by rising client demand. Q2 marked the holdco’s 21st consecutive quarter of growth, with net revenues up 4.8 percent year-on-year – and Publicis raised its full-year organic growth guidance to 4.5 percent–5 percent, lifting the bottom of its previous 4 percent–5 percent range.
Skechers Appoints Horizon Global Across 31 International Markets
Skechers has named Horizon Global — the joint venture between Havas Media Network and Horizon Media — as a global media partner following a competitive review, with a remit spanning 31 markets, Marketing Dive reported. The footwear marketer is focused on tapping into BluConverged, an AI-powered offering that brings together audience strategy, data, activation and reporting, as it looks to build out its omnichannel capabilities and accelerate international growth.
Visa: Stablecoins Will Power Micro-Commerce in the Agentic Economy
A new report from Visa and analytics firm Artemis found that traditional card rails will primarily handle macro-commerce in the agentic economy, while stablecoins – a type of cryptocurrency – will be used for micro-commerce – frequent sub-dollar payments between software systems for services such as API calls or compute resources — with Visa expecting agentic commerce to adopt a hybrid approach combining both rails at different stages of a task.
Media
George Lucas Says AI Sceptics Are Like People Clinging to “Horses and Carts”
Speaking to promote the upcoming opening of the Lucas Museum of Narrative Art, Star Wars creator George Lucas said: “Artificial intelligence means it’s much easier for us to make movies. It’s very much like sitting here saying, ‘Well, I believe the horse and the buggy is really where it’s at.’ There’s nothing you can do about it. That’s progress. It’s the future.” Lucas also argued that AI is better placed than humans to distinguish authentic from fake content, and reaffirmed that individual responsibility and not platform restriction should be the governing principle of AI use.
Netflix Used GenAI in Around 300 of Its Films and Shows This Year
Roughly 300 Netflix titles have now used GenAI in post-production, disclosed in the company’s Q2 2026 earnings report, with co-CEO Ted Sarandos positioning the tools as delivering results that are “10 percent better” creatively rather than leading with cost savings. Netflix said AI is helping its teams work faster without making movies and shows feel smaller or less ambitious, noting that some large-scale scenes may never have been finished without GenAI.
Germany’s Media Regulator Rules AI Overviews Are Subject to Media Law
Reuters reported this week that Germany’s media regulator has ruled that Google’s AI Overviews fall under German media law – one of the first formal regulatory determinations in Europe that AI-generated search summaries constitute a form of media output subject to editorial accountability standards, with significant implications for how other EU member states may approach AI in search.
Publishers Sue Google Over Gemini AI Training
The Guardian reported that a group of major book publishers has filed a lawsuit against Google, alleging that its Gemini AI models were trained on copyrighted content without authorisation or compensation. The case is one of several active lawsuits now running in parallel across the US and Europe targeting the major AI labs, and puts additional pressure on Google at a moment when it is also facing CMA scrutiny and publisher opt-out demands in the UK.
Cox Media Group Taps Waymark for AI Ad Creation
US media conglomerate Cox Media Group has partnered with AI video creation platform Waymark to bring AI-powered ad creation tools to its local advertising clients, allowing small and medium-sized businesses to produce broadcast-quality video ads faster and at lower cost than traditional production workflows.
Tech
GumGum Launches Mindset Agent to Automate Brief-to-Segment Activation
AI and digital advertising company GumGum has launched Mindset Agent, an AI agent that takes an advertiser’s brief and automatically identifies the most relevant contextual segments for activation, removing a manual step from the planning workflow and enabling faster, more precise contextual targeting at scale.
Google’s AI Image Editor Pics Gets an Official Rollout Date
TechRadar reported this week that Google has confirmed a launch timeline for Pics, its AI-powered image editor that takes direct aim at Canva and Adobe for everyday creative and marketing use cases, integrating generative editing capabilities directly into Google’s productivity ecosystem.
Little Dot Studios Launches HyperLaunch, an AI-Powered YouTube Scaling Tool
Little Dot Studios has launched HyperLaunch, an AI-powered tool designed to accelerate YouTube channel growth by automating elements of content strategy, thumbnail testing, metadata optimisation and publishing workflow – targeting media owners and brands looking to scale YouTube presence without proportionally increasing headcount.
Startup of the Week
Reelful is an AI-powered video creation app designed for founders, business owners and creators who want to post consistently on social media but don’t have the time or skills for traditional video editing. The app launched on iOS this week. Users enter a text prompt describing the story they want to tell, record a 30-second voice sample to create a voice clone, then select photos and clips from their camera roll — Reelful plans the video, writes the script, and adds an AI voiceover, assembling the final edit with captions, music and sound effects. The app was founded by Kate Deyneka, a former machine learning engineer at Snapchat.
Number of the Week
The gap between volume and value of AI-powered content is the most honest summary of where the industry sits right now. AI has made it dramatically easier to produce more, while the harder problem of whether any of it is better remains largely unsolved.
