The market is shifting in real time. This time last year, the word “agentic” might not have registered with the majority of marketers. In the last twelve months, AI agents and large language models (LLMs) have moved from theoretical aids to part of everyday life, increasingly becoming extensions of our human selves. They’re actively researching products, logging into accounts, and even making purchasing decisions on behalf of consumers.

If we take a step back, this fundamentally changes the game in terms of marketing. Looking back 15 years ago, we were fighting the battle of bot or not, considering automation as a simple binary decision. If it wasn’t a human, it wasn’t your intended audience. 

But today, the lines are a lot blurrier: the game is much more about trust and intent than “automated = bad.” Nowhere is this more critical than on brand properties.  For marketing teams, this represents a fundamental change in how digital traffic behaves, how it impacts measurement, and ultimately, how it shapes business decisions.We’ve tracked a 6,900% increase in agentic AI activity over the past year. That’s a significant increase from the 1,300% rise we initially reported in October. Consumer adoption tells the same story: 64% of consumers plan to use AI for holiday shopping in 2025, up from just 11% in 2024. Nearly three in four consumers are willing to consider using AI-powered browsers that make purchases on their behalf. The shift isn’t coming. It’s here.

Automation in Media Isn’t New, But AI Agents Are Revolutionizing the Consumer Experience

Automation in digital advertising isn’t new. Invalid traffic, bot networks, and synthetic activity have plagued the ecosystem for years. HUMAN was built specifically to detect these forms of automation in media; our original purpose was to identify and classify the non-human traffic that skews campaign performance and wastes media budgets.

Agentic traffic is just another form of automation we detect. But unlike the bots and IVT of the past, AI agents are active and intentional. These are systems designed to behave like users, navigate digital experiences, research and compare options, and extract the information they need to make decisions. They log into accounts, add items to carts, browse product catalogs, and trigger the same systems and workflows that human users do.

This fundamentally changes the consumer journey.  AI agents are now part of your target audience, even if indirectly. They convert. They matter. 

This means agentic traffic cannot be thought of as a binary –  good vs. bad, bot or not. Automation is no longer a proxy for whether the traffic is malicious or not. There’s nuance, and there’s a need to understand how it behaves, just like how marketers understand the behavior of their human audience. What’s needed to support agentic commerce looks different than what the average user needs. Marketing teams need visibility into agent behavior and intent to optimize the full consumer journey and drive business results. 

How AI Agents Change Everything

When agents log into consumer accounts, research products, and convert, those events get recorded. An agent that browses inventory becomes part of your business intelligence. And they’re now part of your addressable market. 

But you can’t optimize for agents, or understand how they differ from humans, if you can’t delineate between different types of traffic: humans, bots, or AI agents. 

Marketing teams today lack visibility into agentic activity on their sites, which means they can’t see how agents interact with their properties, what they prioritize, or how they influence purchasing decisions. In fact, HUMAN’s recent Iris Report shows that less than half of senior marketers surveyed feel confident their organizations can accurately distinguish between human, bot, and agentic traffic. Understanding that distinction is essential for making informed decisions about how to serve both segments while identifying and detargeting the malicious bots.

For brands, this is a critical juncture: agents are now part of your target audience, but you lack visibility into how they behave, what they need, or how they influence—or, indeed, complete—purchasing decisions. Understanding how agents interact with your properties, what they research, what they convert on, what they need, is what enables agentic commerce. With that visibility, you can customize web experiences for different traffic types: streamlined journeys for agents, optimized experiences for humans.

HUMAN Was Built to Detect Automation, Autonomous or Not

HUMAN was built to identify and classify automation in digital advertising. This includes bot networks, crawlers, LLMs and agentic traffic at both the campaign level and on-site. At the campaign level, we detect agentic traffic the same way we handle all automation. 

But on-site is different; unlike the traditional marketing funnel created for humans, agents can bypass the ad ecosystem entirely. That’s where they’re causing unmeasured impact, a challenge that campaign-level detection doesn’t address. That’s where we identified the gap, and why we prioritized building visibility for brand-owned web properties. The question of whether to block becomes more nuanced, and instead, there’s a need to understand and engage with agents strategically. 

On-site is where agents are actively going to research products and interact with brands on behalf of consumers. For brands, this is where the real impact happens. Marketing teams need visibility on-site to understand this traffic segment separately and make informed decisions about how to engage with agentic commerce. This understanding is what drives smarter marketing and business decisions.

Whether it’s Perplexity Comet, ChatGPT Atlas, or the next agentic tool, each new platform makes AI more accessible to more people. The scale of adoption is increasing as consumers become more comfortable leveraging these solutions. Without this clarity, brands lack the insight to optimize marketing and commerce experiences.

HUMAN’s AgenticTrust shines a light on the AI agents visiting your website. It gives brands direct visibility into agentic activity hitting their sites in real time: understanding agent behavior and intent, identifying which conversions came from agents versus humans, and setting agent-specific controls that enable brands to safely tap into agentic commerce by choosing what actions agents can and cannot take based on what’s best for their business.

In parallel with AgenticTrust, we’re providing granular insights specifically into agentic traffic patterns at the campaign level through dedicated signatures from popular AI and LLM tools, giving customers campaign-level visibility into agent behavior. We’re continuing to evolve our IVT taxonomy to break out more granularly these emerging automation frameworks, which we already identify as automatic browsing, so our customers have increasingly nuanced visibility into agent behavior and composition within their campaigns and supply.

Turning Visibility Into Action

Once brands are able to see how agents are interacting with their properties, decisions become clear. Below is an example of how a brand may use insight from AgenticTrust to inform business decisions and outcomes. 

The Path Forward: Thinking Differently About Agentic Traffic

As agentic AI becomes more sophisticated and agent-based traffic continues to grow, marketing teams will face a choice: treat agents as noise to block, or treat them as an audience segment worth understanding and engaging with strategically.

The brands and marketing organizations that move first will be those that gain clarity on what’s actually happening. They’ll understand agent behavior separate from human behavior, recognize agents as part of their addressable market, optimize dedicated flows for them to serve agent needs while maintaining clear visibility into separate segments. They’ll set baselines and measure performance accurately for both human and agentic traffic, building business strategies that account for both segments.  They’ll be able to justify their marketing and commerce decisions with confidence, and ultimately they’ll drive better consumer experiences and measurable business results.

For everyone else, the noise will only get louder.

If you want to see what this clarity looks like in practice, request a demo. We’ll show how agentic traffic can be identified, segmented, and measured separately from human traffic, so you can understand what’s actually happening in your funnel and make decisions with evidence instead of assumptions.