Futurecasting can be difficult when technology changes on a daily basis. But we asked some of the brightest minds within The Marketing Cloud (and the broader Stagwell network) to weigh in on what the field of agentic AI might look like in the coming year.
As companies compete to incorporate AI into their business, those who come out as winners in 2026 and beyond will be those who embrace these agents as brand ambassadors. AI is already evolving SEO and will eventually overtake how consumers learn and hear about brands. No longer will companies be able to control their brand narratives and PR in the way they do today. Brands who actively shape agents and innovate across every aspect of their business will be able to adapt to this wave of technology.
At the same time, as marketing grows more efficient, marketing budgets will begin to expand. Agentic AI will make everything from consumer interface to production of content significantly easier and less expensive. This means clients will spend more because they get better returns on their investment. Just as brands needed digital transformation experts to build websites and apps, they now need them to build agentic AI. There will be a whole new round of agentic AI transformation work, ballooning marketing budgets in 2026.”
—Mark Penn, Chairman and CEO of Stagwell
However, similar to human teams, a shift from generalists to specialists is necessary. Specialists enable deep focus, technical expertise, and the competency to tackle more intricate problems.
In 2026, I anticipate AI agents starting to emerge with specializations in methodologies, quality assurance, analytics, and reporting. This evolution will further boost our ‘Insight Advocates,’ experienced researchers who link AI-derived insights with organizations to address business-related questions.
In practice, that means the AI agents will operate independently across the research process while maintaining a human 'in-the-loop' to ensure ethical standards, governance, and human oversight are upheld.”
—Gary Topiol, Managing Director, QuestDIY
By embedding generative AI, 24/7 intelligent agents, and predictive analytics across the marketing lifecycle, we can unlock significant improvements in performance metrics, drive significant impact, and outpace less innovative competitors.
This isn’t just about automation. It’s about rearchitecting marketing as a dynamic, data-driven engine for strategic advantage.”
—John Kahan, Chief AI Officer, Stagwell
It’ll happen once we can connect disparate consumer data sources—from purchase history and loyalty program behaviors to motivations, affinities, engagement, customer service history, and more—with AI agents that have the creative and content, rules-based logic and ability to activate channels and make decisions that increase a consumer’s Life Time Value in real-time.
This future enables brands to truly learn from and adapt to consumer behavior, while at the same time breaking down silos that can exist across various channel teams that want to control bits and pieces of the consumer journey. Now the consumer—and their behaviors across every touchpoint—can drive the experience instead, and it’s up to brands to continually optimize.”
—Elspeth Rollert, CEO, The Marketing Cloud
What might this look like in practice? A powerful LLM will act as the strategist, breaking down a task into smaller problems, where the moment-to-moment execution will be handled by fleets of hyper-efficient Small Language Models (SLMs).
These discrete steps or problems—such as personalizing thousands of ad creatives, A/B testing subject lines or creatives in real time, or generating on-brand video clips—will be handled by these specialized models. This will transform the human marketer from a creator to a conductor of an ‘AI orchestra,’ fine-tuning individual SLMs for specific tasks to deliver campaigns that are not only adaptive and personalized but economically scalable.”
—Eric Walzthöny Kreutzberg, co-founder and CTO of SmartAssets
A very big problem for brands is that these AI browsers can function like sales reps from other companies stationed in your store. When a customer lands on your product page, agentic AI could instantly contextualize and present your competitors, side-by-side. The experience you painstakingly designed becomes commoditized into a generic comparison feed.
This accelerates a dangerous trend: the erosion of brand ownership. As agentic systems mature, customer acquisition costs will rise, loyalty will decline, and brands without strong first-party data will find themselves struggling for differentiation.
The good news: The technology isn’t there, yet. Today’s AI browsers excel at ‘better thinking,’ not ‘better doing.’ They summarize, reorganize and compare, but they don’t execute complex, multi-step tasks.
That lag gives brands time to innovate where Google cannot: industry-specific, convergent user experiences that can anticipate customer needs before a copilot intervenes.
In 2026, marketers who seize this window will have the chance to create irreplaceable experiences. Those who don’t will wake up to find Google, Perplexity and others have become the real point of contact with their customers.”
—Dan Gardner, founder and CEO of Code and Theory (part of the Stagwell challenger network)
In 2026, this won’t replace proven methods; it will amplify them. Agents will streamline how data is connected, analyzed, and highlighted, giving Insights and Brand teams sharper, cleaner outputs on demand. With QuestBrand, we are actively testing and building these capabilities to compound the value of traditional brand equity measures with agentic intelligence. The pace of the market demands nothing less.”
—Zeke Hughes, Managing Director, QuestBrand
In 2026, what we want an agent to do and what it is capable of doing will continue evolving daily. There won’t be one kind of agent, but rather an assortment. Slow and fast, simple and complex, this variety of agents will all ‘live’ and thrive in a sort of agentic hive.
Think of multiple marketing agents, for instance: Creative agents, Media agents, Optimization agents, and Reporting agents all working in cohesion within an agentic framework.
That agentic framework itself is also in flux. Sure, everyone would like to see a fully autonomous agent. But much like a fully self-driving car, this tech will initially have drawbacks, limitations, and regulations to abide by at first. We should be cautiously optimistic as we enter this exciting new world of AI and marketing.”
—Mansoor Basha, CTO of The Marketing Cloud