Is AI talking about you behind your back? A model for the agentic age

For more than 50 years, marketing has been obsessed with architecting the perfect journey. We’ve mapped every touchpoint, sequenced every message and optimized every micro-moment. From awareness to consideration to conversion, the funnel became doctrine.

It’s no surprise, then, that much of today’s conversation about AI in marketing is tactical: How can we use it to produce assets faster? To optimize creative? To personalize at scale? To improve efficiency across journeys we already know? That’s a logical place to start. But a bigger question is emerging, one that sits upstream of creative production and performance marketing: What happens when AI mediates discovery, evaluation and how consumers choose brands?

This isn’t a future scenario. It’s happening now. For many consumers, AI has become the middleman you didn’t hire. For these consumers, AI isn’t a new channel. It’s the intermediary through which all channels are filtered, synthesizing and reconstructing brand meaning.

When a user asks Perplexity, ChatGPT, Gemini or Claude for a recommendation, research and evaluation collapse into the model itself. AI does not experience your brand through your funnel. It doesn’t navigate your site sequentially, read case studies in order or absorb positioning through carefully crafted messaging. Instead, it synthesizes what it knows—scraped from the web, structured from databases, inferred from training data and contextualized against competitors. It provides a judgment in seconds. Your brand is no longer encountered as a series of messages. It is encountered as a conclusion.


is-ai-talking-behind-your-back-editorial1




The defensive reflex

The initial response to this shift has been defensive. Brands are rushing to optimize websites for agent crawlers, structure content for clean scraping and hope models cite them accurately. These efforts are necessary but not sufficient.

Agents don’t simply read content. Increasingly, they reference one another, exchange information and develop shared understandings. If a brand exists only as static content, it will be reduced to an inferred summary: a passive source rather than an active participant. Emerging protocols such as MCP are establishing how agents will communicate functionally, structured channels brands will need to navigate.

is-ai-talking-behind-your-back-editorial2




The structural problem with the modern website

This exposes a structural problem with most websites. Built for an SEO paradigm now being deprecated, they are not prepared for agentic consumption. Content is largely static and difficult for agents to parse. Headlines and hero images dominate, while verifiable facts and decision-relevant context are fragmented or buried.

When agents scrape these environments, they extract fragments without nuance. The result: brand meaning reconstructed with incomplete information.

is-ai-talking-behind-your-back-editorial3




From static sites to intelligent brand systems

This is where generative interfaces become interesting. Instead of treating websites as static repositories, brands can create generative, conversational layers that serve both humans and agents. These systems generate content from structured knowledge in consultable ways, with clear sourcing and real-time adaptation. For humans, these interfaces feel natural and responsive, adapting dynamically to what a visitor is trying to understand or achieve. For agents, they provide clean structure, verifiable data and the ability to resolve ambiguity. One system, two audiences. An intelligent brand system.

This transition brings discomfort. Brands can no longer script every interaction when content is generated in real time. What they can control is the intelligence the system draws from: defining the knowledge base agents consult, establishing guardrails around what can be said and shaping how uncertainty or confidence is expressed. Semantic foundations and ontological frameworks become core brand assets.


is-ai-talking-behind-your-back-editorial4




Rethinking measurement in an AI-mediated world

This shift demands new measurement. Traditional metrics—reach, impressions, click-through rates—focus on visibility, which matters less when AI mediates interaction. More meaningful questions emerge:

- How does an AI system describe your brand?
- Which competitors are mentioned alongside you, and why?
- What attributes or weaknesses are consistently surfaced?
- How stable are those answers across models and over time?

These signals are measurable, even if not standardized. McKinsey reports that only 16 percent of brands are actively tracking how AI represents them. That gap represents opportunity. Authority within intelligent systems is still being established, and early movers can help define it.


is-ai-talking-behind-your-back-editorial5




From attention to authority

The brands building generative and agentic experiences aren’t simply applying AI to existing tactics. They are rethinking how value is created and demonstrated when intelligence itself becomes the interface.

In this landscape, attention is no longer the scarce resource. Authority is. And the brands that learn to participate in the conversation—rather than simply publish and be summarized—will shape the intelligence age.

R/GA, Marketing Agency

Is AI talking about you behind your back? A model for the agentic age

The

defensive

reflex

The

structural

problem

with

the

modern

website

From

static

sites

to

intelligent

brand

systems

Rethinking

measurement

in

an

AI-mediated

world

From

attention

to

authority