Loading...
Four signals landed in one window: a holdco claimed the FDE category, a platform auto-ran creative, layoffs deepened, and two-thirds of CMOs called their answer shaky.
In the last twenty-four hours, four signals landed in the same window. Together, they redraw the line between who is doing AI work inside a brand and who is left defending what got done. A major advertising holdco publicly claimed the forward-deployed engineer category for itself. A major platform's AI ad tools shipped a bike with two handlebars for one of its oldest retail customers. A major holdco announced hundreds more layoffs across its flagship creative network. And two-thirds of chief marketing officers at a closed-door summit admitted, when polled live, that their honest answer to "what did marketing deliver" is "shaky." The FDE is now publicly staffed. The CMO's seat is publicly buckling. The strategy-accountability layer above both is the open one.
On July 14, 2026, WPP launched Hex — its own forward-deployed engineering studio. The Ad Age report describes Hex's FDEs as "responsible for providing vital AI expertise to organizations in an up-close-and-personal way." It is the first major advertising holding company to publicly claim the FDE category as its own, putting the holdco directly alongside the AI labs, the Indian IT services tier, and the consulting firms that have spent the last year staffing the same lane.
Hours later, the same outlet reported that WPP is cutting hundreds more jobs by the end of 2026, with VML among the affected networks. The restructuring under new CEO Cindy Rose is no longer confined to media-buying or back-office functions. It now reaches the named creative flagship.
The same day, Business Insider reported that REI ran an Instagram ad depicting a bicycle with two sets of handlebars. The retailer said Meta had auto-enrolled it in an AI creative feature that produced "inaccurate" and "inappropriate" imagery. A Meta spokesperson responded that "AI can make mistakes and that it is the advertiser's responsibility to review the AI outputs." The accountability hand-off — from platform to brand — is now public, named, and on a bike.
On July 13, a separate piece in MediaPost reported that two-thirds of CMOs at a closed-door summit, when asked live how good their honest answer would be to "what did marketing delivered," said "shaky." The piece, titled "Confessions of an Agency CEO: We're Overselling Fries," lands the same week the holdco is laying off strategists and the platform is auto-enrolling advertisers in features that reshape their product photography.
Three of these four signals are on the supply side. The holdco has FDEs on the ground. The platform is auto-running creative. The restructure is shaving cost out of the named creative network. The fourth signal is on the demand side — and it is the one that should stop every CMO in the room.
When the people who execute AI work are increasingly well-staffed, and the people who defend what AI produced are increasingly few, the gap between deployment and accountability widens. A platform can say the brand is responsible for reviewing AI outputs. A holdco can say the FDE is in the room. Neither answer resolves the underlying question: who owns the strategic decision, defends the recommendation in the boardroom, and absorbs the consequence when the model — or the agency — gets it wrong.
This is the named-counsel layer. It is the seat that decides what to deploy against, what to approve before it ships, and what to roll back when it fails. It is the layer above the FDE, above the platform's auto-enroll, and above the holdco's restructure.
Until this week, the FDE lane was publicly occupied by AI labs, the largest services firms, and a handful of consulting-grade agencies. Tata Consultancy Services publicly committed to up to 8,900 forward-deployed engineers for AI deployment. TCS then signed a multi-million, multi-year deal with ABB to operate ABB's global network as a service, with AI embedded — a 20-year partnership extension. Tomasz Tunguz tallied roughly $9.75 billion in 12-month FDE commitments across OpenAI, Microsoft, Anthropic, Amazon, and Google Cloud. The deployment tier was never going to stay unclaimed by a major holdco for long. Hex is the claim.
The week of these announcements also saw Meta nearly double the projected cost of its Louisiana AI data center to more than $50 billion for 5 gigawatts of compute capacity. The infrastructure race is not slowing. The deployment race is not slowing. The only question is which layer above them is staffed, accountable, and capable of defending a decision when the model, the platform, or the holdco gets it wrong.
Two-thirds of CMOs admitting their honest answer is "shaky" is not a question of how much AI they have bought. It is a question of whether the work product coming out of the AI is something they can defend in the room where the budget is decided. When REI's bike had two handlebars, the CMO was the one answering for it. When the holdco lays off the named network and the platform auto-enrolls the brand in creative features, the CMO is the one explaining the gap. When the FDE is in the room but the recommendation is not, the CMO is the one whose seat is at risk.
The named-counsel layer is the layer that absorbs that risk. It is the layer that names the strategic objective before the FDE deploys against it. It is the layer that owns the recommendation after the platform auto-enrolls the brand. It is the layer that defends the decision in the room where two-thirds of CMOs are already admitting their answer is shaky.
A year ago, the brief for an AI engagement was: what tool do we buy, what data do we feed it, what dashboard do we read. The FDE boom changed the brief to: who do we put in the room to deploy against the strategy. The holdco's claim, the platform's auto-enroll, the holdco's restructure, and the CMO's buckling all change the brief again.
The new brief is: who owns the strategy while the FDEs ship, while the platforms auto-run, while the holdcos restructure, and while the CMO is left defending the result. The named-counsel layer is the answer to that brief. It is the layer above the FDE, above the platform, above the holdco, and above the tooling tier. It is the only layer that does not move when the infrastructure, the deployment model, or the holding-company chart changes.
Autostrat is the named-counsel layer. We do not deploy engineers into the room. We do not run a platform. We do not operate a holdco. We deliver the strategic decision — the named objective, the recommendation, the defensible position — that the FDEs ship against, that the platform runs against, that the holdco restructures around, and that the CMO defends in the room. One subscription replaces the question of who owns the strategy while every other layer moves.
The FDE stood up this week. The CMO's seat buckled the same week. The layer above both is open, and it is the one that decides whether the next twelve months of AI work produces decisions a board can defend or another quarter of "shaky" answers the CMO has to walk into the room with.
Book a 30-minute demo. Bring a live question and watch the answer get built.