1 Warner Bros. Discovery is rebuilding its entire ad stack on agents
Warner Bros. Discovery said it's rebuilding its whole advertising technology stack around agentic AI on AWS — automating media planning, audience forecasting, measurement, attribution, order management, pricing and campaign stewardship. It joins WPP, Dentsu, Dept and others expanding agentic offerings, but a media owner re-platforming its own ad business is a bigger tell than another agency press release.
When the sell side rebuilds itself around agents, the buy side has to follow or get optimized against. The practical question for any marketer running CTV or premium video spend: who is the agent on the other side of the table working for? Start asking your media partners what's automated, what's auditable, and where a human still signs off — because "agentic" on the supply side quietly changes how your money gets priced.
2 Microsoft put an agent inside the CRM
Microsoft's Journey Creation Agent for Dynamics 365 Customer Insights went live in North America and Oceania, building multi-step customer journeys from a natural-language prompt instead of a manual canvas. The job that used to mean dragging boxes around a flowchart is becoming a job of reviewing what the agent drafted.
This is the quiet automation that actually changes headcount math. The skill that gets valuable isn't building the journey — it's knowing which journey is right, catching where the agent's logic is wrong, and owning the strategy the prompt can't infer. If your team's CRM value is "we can operate the tool," that moat is closing. Move people toward judgment, segmentation strategy, and offer design now.
3 Hyundai built its own AI media buyer
Hyundai stood up an in-house AI-powered media-buying system that, by its own account, reduced costs and improved campaign performance. It's the clearest version of a trend hiding under all the vendor noise: big brands aren't just buying agents, they're building them — and pulling capability back in-house.
The strategic read for everyone smaller: you won't out-build Hyundai's data science team, so your edge is the opposite. Lean into the things in-housed automation is bad at — distinctive creative, first-party relationships, category point of view — and use off-the-shelf AI for the media mechanics. The brands that lose this cycle are the ones that try to half-build what they can't fund and half-outsource what they should own.
4 Asda handed its retail media to Amazon
UK grocer Asda signed a deal to deploy Amazon's Retail Ad Service across its online grocery platform and its George fashion and home label, rolling out in stages from Q4. It's a tell about where retail media is heading: even large retailers would rather rent Amazon's ad tech than build their own, as US retail media spend climbs toward $71.09 billion in 2026, up 17.8% year over year.
For brands, the consolidation cuts both ways. Buying gets simpler when more networks run on shared infrastructure — but it also means Amazon's measurement, taxonomy and rules increasingly set the terms everywhere, not just on Amazon. Audit how much of your "retail media" budget is quietly becoming Amazon-flavored, and make sure you still control your own first-party data and creative.
5 A $1.1M pre-seed bet that AI chat is the new shelf
Nudge raised $1.1 million in pre-seed funding and launched an Agentic Commerce Platform to track and manage how products get recommended across AI chat apps. Translation: someone is building the "rank tracker" for the era when ChatGPT, Gemini and Perplexity decide which brands get mentioned in an answer.
Tiny round, big tell. If shoppers increasingly ask an assistant before they ever hit your site, then "what the AI says about us" becomes a measurable, optimizable channel — answer-engine optimization with a dashboard. You don't need to buy a tool yet, but you do need to start checking what the major assistants actually say when asked to recommend products in your category. That answer is your new homepage above the fold.
6 AI shopping grew 393% — and still moves under 1% of traffic
Here's the reality check the keynotes skip: AI-sourced retail traffic grew 393% year over year in early 2026 and converts better than non-AI traffic — yet it still sits under 1% of total retail traffic. Google still owns 56.68% of shopping search starts, Amazon 28.96%, and ChatGPT just 7.26%.
Both numbers are true and both matter. The growth rate says prepare now: make your product data, reviews and pages legible to agents, because the curve is steep. The base rate says don't reallocate your whole budget chasing 1%. Treat agentic commerce like a 2027 channel you're funding the foundation for in 2026 — enough to learn, not so much that you starve the channels paying today's bills.
7 Impact bought ACTIVATE — influencer marketing keeps consolidating
Impact.com acquired ACTIVATE, an end-to-end influencer marketing platform, folding creator discovery, campaign management and reporting into its Partnership Cloud. It's the latest in a steady run of consolidation that's pulling influencer tooling under platforms built to connect creator activity to verified transactions.
The pattern is the point: the capital and the acquisitions are flowing to whoever can tie a creator post to a tracked sale. If your influencer program still lives in spreadsheets and screenshots, you're running the version of this channel that's being acquired out of existence. Push your partners — or your stack — toward real attribution, because "engagement" is no longer the thing buyers are paying for.
8 Havas paid up for human craft — and bought Archrival
While everyone else bought agents, Havas bought people: it acquired youth-culture and experiential agency Archrival to grow Havas Play in North America. In a month where AI flooded the production pipeline, a holding company spent real money on hard-to-automate cultural fluency and live experience.
Read it next to the agent launches and the story sharpens. The work that's getting cheaper — variations, resizes, drafts — is racing toward zero. The work that's getting more valuable — knowing what's cool, building something people show up for in person — is what the holdcos are acquiring. For brands, that's the budget signal: let AI handle volume, and protect (or hire) the humans who make the thing worth talking about.
9 The creator economy got bigger and harder to fund at the same time
The US creator economy is on track to clear $40 billion in 2026, but venture funding for creator startups cooled hard — roughly $58 million across nine deals in the first five months, versus about $807 million across eleven deals a year earlier. The money that did move concentrated in creator-marketing infrastructure like Devotion, Humanz and Agentio.
The split is the story. Investors stopped funding "creator" as a vibe and started funding the plumbing that makes creator spend measurable and scalable. For brands, that mirrors where your own dollars should go — less on one-off sponsorships, more on the systems that let you run creators like a performance channel with real attribution. The creator-as-awareness-flyer era is closing; the creator-as-tracked-line-item era is here.
10 Optimizely rebranded around the "AI does the mundane" promise
Optimizely unveiled a new brand identity, with CEO Alex Atzberger framing it around the idea that AI should make the mundane parts of marketing easier so brands can focus on human curiosity, creativity and compassion. It joins a crowded month of martech players repositioning as AI-native.
Watch the framing, not just the logo. "AI for the boring parts, humans for the meaning" is becoming the consensus pitch — and after the month we just had, it's clearly the right one — but it's now table stakes, not a differentiator. When every vendor says the same sentence, ignore the rebrand and test the actual workflow: does the tool genuinely remove a chore your team hates, or just add an AI label to the dashboard?