Social Rocket
DMEXCO 2026 has moved the AI debate in marketing one step further: the focus is no longer on the next pilot, but on how companies can bring artificial intelligence into everyday work safely, measurably and in line with their brand. On panels with leaders from Meta, Gamma, Taxfix, IBM, Mini, Decathlon and Lorenz Snacks, it became clear that scaling is now primarily an organisational and leadership task.
This matches the fair's official motto, "Scaling Intelligence". Before the event, DMEXCO had announced that it would focus on the transition from AI experiments to concrete processes and business models. According to the organiser, more than 40.000 visitors came to Cologne.
Governance before speed
In the discussion on AI scaling documented by Digiday, governance and trust moved conspicuously to the front. Asil Yilmaz of IBM pointed out that companies should review internal rules and possible gaps before rolling out new models widely. Tino Krause of Meta stressed the importance of employees who can question AI results rather than follow them automatically.
For marketing teams, this is a practical shift: access to a model is no longer necessarily the bottleneck. What matters are approvals, responsibilities, data access and which tasks automation is actually sensible for.
This complements developments in media buying. A recent IAB Europe survey shows how quickly agentic media buying is moving into day-to-day operations. At the same time, the need for reliable control mechanisms is growing.
Don't automate every task at once
On the panel, Alexander Beresford of Taxfix argued for looking carefully at tasks, responsibilities and goals rather than deploying AI everywhere. Gamma CGO Kristin Fracchia, meanwhile, described the high pace: what teams can implement with AI today in a few months would previously sometimes have taken considerably longer.
That is where the tension lies. Faster production is not yet better marketing. Anyone scaling AI must therefore measure not only time savings, but also quality, brand impact and actual business value.
Measurement also creates new challenges. When AI systems themselves become research, recommendation and purchasing interfaces, a single visibility score is not enough. Our overview of measuring AI visibility explains why sources, citations, referral traffic and conversions should be considered separately.
Brand leadership remains human
On the second day of the fair, the discussion shifted more towards brand leadership. Representatives of Mini, Decathlon and Lorenz Snacks described AI as an efficiency lever, but also warned against making brands interchangeable through uniform, data-driven communication.
This is especially relevant for social and content teams. Automation can speed up research, variant creation and production. However, brand identity, cultural sensitivity and the final decision on what is actually published cannot simply be delegated.
In practice, this argues for workflows with clear brand rules and human approval. Our guide to using Claude as an AI creative director shows what such an approach can look like in concrete terms.
What marketers can take from DMEXCO
The central development at DMEXCO 2026 is less a new AI tool than a process of maturing: companies must decide which tasks to automate, how to safeguard quality and which decisions deliberately remain with people.
In parallel, AI is also changing the customer journey. Brands are already observing how AI search is influencing media planning and upper-funnel budgets. This makes AI governance, brand leadership and measurement increasingly parts of the same marketing strategy.
Sources: Digiday: DMEXCO ’26 Briefing, DMEXCO: Highlights 2026, DMEXCO CMO Summit