The agentic eagle has landed. LLM crawlers and AI agents are beginning to transform how consumers, businesses, and fraudsters interact online. The most notable metric is the explosion of LLM/AI traffic now hitting websites, mobile apps and APIs. DataDome customers, for example, saw AI/LLM request traffic almost triple between January and May 2025, from 2.6 to 7.2% of total bot traffic. 

Why agentic AI traffic matters for your business

Such traffic, however, is just a leading indicator of the sea change in how online businesses will operate. Unfortunately, agentic AI generally means less visibility for marketing, revenue, and security teams, resulting in:

• Negative SEO impacts and dramatic traffic drops
• Less session & user information to build customer relationships 
• Disruption of sales (and thus revenue) opportunities
• Business logic abuse that’s harder to block with autonomous bypass activity, ranging from credential stuffing to cart manipulation attacks
  

Without visibility into the actual agentic AI and LLM crawler site activity, many teams currently have one arm tied behind their backs. There’s a blurring between legitimate and illegitimate users and the request intent.

And we’re likely at the beginning of the proverbial hockey stick growth curve, since more enterprise software applications will include agentic AI. Gartner predicts that 33% of such apps will include agentic AI by 2028.

DataDome brings visibility & control

That changes with DataDome. Our mission is to give you visibility and control over AI traffic. A couple quarters ago, DataDome began to classify this new traffic type explicitly, showing agentic AI and LLM crawler traffic directly in our dashboard.  That was based on detection advancements around LLM and AI traffic that our threat research experts have provided for more than a year. This provides the following benefits instantly:

• Groups AI/LLM traffic into a dedicated category automatically 
• Shows what is accessing digital assets and why
• Provides intelligent policy recommendations

So far, this has helped security teams at our customers to quickly understand and respond to this traffic based on bot identity, behavior, and trustworthiness. We’ve even compiled and published information on the most common AI agents & LLM crawlers publicly.

What’s new in the threat dashboard

Now, DataDome has enhanced our threat dashboard further to make it easier and faster to see and secure your sites for agentic AI & LLM crawlers. Among the new capabilities are:

• A new side menu category listing with a dedicated explore filter.
• Far more user, session, and intent-based data to inform and characterize agentic traffic across protected endpoints.
• More reporting capabilities to share information across teams.

The first step to security for your digital properties is visibility. Now you have that in spades. If you want to learn more, we’d be glad to give you a demo. Or, We’re hosting a 3-hour workshop next month at MRC San Diego that will give you the clarity and strategies you need to adapt, before agentic AI impacts your revenue and risk.