1 Walmart called itself "AI native" — and Sparky has the receipts
On Walmart's Q1 earnings call, CEO John Furner said the retailer is "becoming AI native," and pointed to Sparky, its AI shopping agent, for proof. Weekly active users on Sparky more than doubled in a single quarter, units purchased through the agent quadrupled since the last quarter, and customers using Sparky show an average order value roughly 35% higher than non-users. Walmart's global ad revenue also grew 37% in Q1, with AI tools cited as a driver.
This is what "AI native" sounds like when it's real: a CEO citing AOV, basket size and ad revenue on the same line. For brands, two things change. First, the merchandising surface inside Walmart isn't a static PDP anymore — it's a chat that reasons about replenishment, meal plans, and substitutions. Second, Walmart Connect just got a closed-loop story it can sell to advertisers. If you sell on Walmart, your Sparky visibility is the new shelf placement. Treat it like one.
2 Pattern launched Pi — commerce intelligence sold as a product, not a deck
Ecommerce accelerator Pattern launched Pattern Intelligence (Pi), an AI-driven commerce intelligence platform that recommends pricing, content, inventory, and visibility moves across Amazon, Walmart, TikTok Shop, and other marketplaces. It's built on Pattern's commerce dataset and trained for the operational decisions brand managers actually have to make — not the dashboards they have to click through.
The marketplace optimization category used to be a quarterly consulting deck; Pi is the same playbook turned into a SaaS subscription. The unlock is speed: pricing moves and listing tweaks that used to wait on an agency cycle now ship in days. The risk is that everyone optimizing on the same model converges on the same answers, so the differentiator becomes the proprietary input — your first-party data, your supply story, your brand — not the model itself.
3 X shipped Creator Connect — xAI is now the matchmaker
X launched Creator Connect, an AI-powered service that uses xAI to match brands with creators based on campaign goals, audience overlap, and live trend signals on the platform. X runs outreach, content production, and distribution as a managed workflow — closer to an agency engagement than a directory — with the brand keeping approval at each step. Early tests included a premium laptop pairing with tech creators who also follow F1, and a horror-film studio matching with niche creators.
X is positioning Creator Connect as the commercial product of its self-declared "Year of the Creator," and inserting xAI in the middle of brand demand and creator supply. For marketers, the appeal is access to micro-creators you'd never find through a marketplace search, paired with real-time trend signals only the platform sees. The catch is that the same AI choosing the creator is owned by the platform selling the media. Audit the recommendations the way you'd audit any house-favored placement.
4 PubMatic put a reasoning agent inside the buying dashboard
PubMatic introduced its Detailed Reasoning Agent as part of its Sage suite, delivering analytical summaries inside programmatic dashboards. The release came with Chat-to-Data for on-demand questions, a Knowledge Management System for brand-specific guidelines, and scorecards tracking how products rank with AI shopping agents like Alexa, Walmart Sparky, and ChatGPT. The pitch: take queries that used to require an analyst, and answer them in the seat.
The deeper move is the AI-shopping-agent scorecards. PubMatic isn't just helping you buy media — it's measuring whether your products show up in the answers consumers are getting from a different generation of intermediaries. That's a new KPI marketers don't yet have a budget line for. They will. Treat agent-visibility as a row in your media plan, even if you're not buying against it yet.
5 DoubleVerify wired itself into Meta Threads
DoubleVerify launched a content verification application for Meta Threads, automatically evaluating brand-adjacent posts at the post level and flagging risk factors using AI. It's the latest brand-safety vendor to formally extend coverage from feed-based social into text-discussion environments — the same surface Meta has been trying to grow as its Reddit-shaped alternative.
Threads is being courted by advertisers exactly because it's quieter, more text-driven, and less algorithmically chaotic than Reels. But "quieter" doesn't mean "safer," and conversations are harder to moderate than static creative. Bringing DV-style verification to that surface is what makes Threads buyable at scale — and it tells you Meta is serious about monetising discussion, not just engagement. If your brand has been allergic to Threads on safety grounds, the rationale just got weaker.
6 Typeform's Growth Flow turned forms into a workflow
Typeform launched Growth Flow, a form-completion system that triggers downstream business actions when respondents finish. Data flows straight to sales pipelines, email lists, and support tickets without manual intervention, and the system uses AI to read sentiment in written feedback and adjust subsequent questions accordingly. It's the dynamic survey grown up into a routing layer.
The interesting bit is the sentiment-driven branching. A form that asks one set of questions to a happy respondent and a different set to a frustrated one is closer to a conversation than a survey — and it generates training data for whatever agentic CX system you build next. If your brand is still treating forms as static gates, you're leaving both data and empathy on the table.
7 Creatify shipped a creative agent built for the algorithm
Creatify launched Creatify Agent, a video production system trained on past ad performance that generates video creative aligned with what's currently working across TikTok, Meta, and YouTube. The pitch: not just "make a video," but "make a video like the ones converting this week on the platform you're buying."
This is the next layer past dynamic creative optimization — an agent that doesn't just remix your assets but ingests the platform's winning patterns and writes back to them. The honest tradeoff: you'll get more competent creative, faster, and a lot of it will look like everyone else's. The advantage is in the input. Feed the agent a real brand voice, not just performance data, or it will optimise you into the average of every brand currently outperforming you.
8 Bayer wrote the rules for agents before they spent the budget
At Digiday's Programmatic Marketing Summit, executives from Bayer and Kelly Scott Madison described the guardrails they're putting around AI agents in programmatic workflows: spend caps, data anonymisation requirements, and internal "gatekeeper" agents that enforce brand standards and campaign rules. The fear is concrete — autonomous systems overspending budgets, misusing data, or making opaque optimization calls — and the answer is policy in code, not policy in PowerPoint.
This is the unglamorous side of agentic marketing finally getting addressed. The right model isn't "trust the agent" or "don't use the agent." It's a gatekeeper agent reviewing the working agent, with humans on exceptions. If you're piloting AI agents anywhere in your buying stack, write the rules now — the spend limits, the data boundaries, the brand-safety lines. The vendor demos won't write them for you, and "we'll figure it out later" is how you end up explaining a $400K test budget to the CFO.
9 Kovva went after the ops layer programmatic forgot
Adtech startup Kovva launched AI agents focused on the repetitive ops work that dominates programmatic management: QA checks, discrepancy investigations, reporting reconciliation, and creative fatigue detection across DSPs and ad servers. Less "make my creative," more "stop the team from doing the same Excel job at 11pm."
The most quietly important agentic launches aren't on the buy side — they're in the back office. Programmatic ops is full of high-leverage, low-glory tasks where an agent can pay for itself in a week. If you run an in-house programmatic team, the question to ask this quarter isn't "what creative can AI make?" It's "what reconciliation can it stop doing by hand?" That's where the headcount math actually changes.
10 Kustomer Architect turned "AI transformation" into a service line
Kustomer announced Kustomer Architect, a capability inside its AI-native CX platform designed to guide brands through AI rollouts toward measurable outcomes — CSAT, retention, operational efficiency, revenue. It's a consulting-flavored layer on top of the product, packaged as software.
Read alongside Bayer's guardrails, this is the maturing of the "we'll AI everything" pitch into something brands can buy with a clear deliverable. The thing to watch isn't the framework — it's whether the metrics it commits to are the ones a CFO already tracks. AI transformation programs that report on CSAT and retention get renewed. The ones reporting on "agent adoption" don't.