This is the year artificial intelligence will evolve from a playground for futuristic ideas into a practical tool driving tangible business outcomes and become an indispensable part of advertising strategies.

Among the myriad tech discussions at CES 2025, agentic AI emerged as the breakout concept, finally getting the attention it deserves.

Agentic AI is more than just another buzzword. Unlike generative AI, which creates content, agentic AI operates autonomously, making decisions and taking actions without human intervention. Nvidia CEO Jensen Huang captured the excitement on the CES stage, declaring agentic AI a “multi-trillion-dollar opportunity.” It’s not hard to see why this concept resonated so strongly for an industry that is always chasing efficiency and innovation. 

The tech that stole the show

Nvidia, in fact, dominated the agentic AI conversation, announcing several innovative initiatives. Its AI Blueprints aim to equip developers with tools to build autonomous systems capable of analyzing data, reasoning and executing tasks. These blueprints integrate seamlessly with the Nvidia AI Enterprise platform, a boon for businesses looking to scale AI adoption. Partnering with Meta, Nvidia also unveiled the Llama Nemotron models, designed for customer service and supply chain optimization. Companies such as SAP and ServiceNow are already lining up to implement these tools.

Meanwhile, Gigibyte introduced its GiMATE AI Agent, an innovation built for next-gen AI PCs. GiMATE promises users smarter hardware-software interaction, enhancing productivity and even gaming experiences. This product aligns with the broader theme of bringing agentic AI into everyday applications.

But Nvidia’s Cosmos platform might be the boldest leap forward. It blends generative AI with advanced video processing for robotics and autonomous vehicles, setting the stage for a future where AI doesn’t just think for itself but acts in the physical world.

A wake-up call for measurement standards

For advertising, agentic AI’s benefits include analyzing data, making decisions and executing tasks that will allow smarter targeting, personalized experiences and streamlined operations without constant human input. 

Yet, our industry faces a challenge it hasn’t fully grappled with yet: how to measure traffic in a world where AI agents are as active online as humans? AI tools that summarize websites or fetch user content might generate impressions, but should those impressions count?

Analytics become unreliable if these visits are treated the same as human ones. A surge in traffic could just as easily reflect a swarm of AI agents as it could genuine consumer interest. For advertisers, this is challenging. Campaign performance metrics could be thrown off, misallocating budgets and eroding trust in data-driven decisions.The issue isn’t just about filtering out bots. Traditional bots often declare themselves, making it easier to exclude them from reports. Agentic AI, however, is designed to mimic human behavior, slipping past filters and muddying the data. Differentiating between authentic human engagement and AI-driven activity is no longer optional—it’s mission-critical.

The edgy truth about AI traffic

AI agents are already reshaping the web, whether we’re ready for it or not. They don’t click on ads. They don’t buy products. Yet, their activity can inflate impression counts, leaving advertisers to foot the bill for interactions that generate zero ROI.

Some platforms are starting to address the problem. Tools such as Nvidia’s AI Blueprints could eventually help distinguish between human and AI traffic, but the industry is far from a consensus on how to standardize these measurements. Until then, advertisers must push for transparency from both AI developers and analytics platforms.

The rise of agentic AI should be a wake-up call, not just for tech companies but for anyone invested in digital advertising. It’s time to rethink what success looks like in this brave new world. If 2025 is truly the year AI becomes indispensable, the industry needs to adapt—fast. Anything less risks turning the promise of agentic AI into a missed opportunity.