Most companies are treating AI agents the same way they treated spam bots a decade ago, and it's costing them billions in lost revenue. Between November 2025 and March 2026, AI agent traffic surged 150% month-over-month, now representing 15% of all website visits. Yet 81% of organizations are still filing these agents under outdated security rules designed for an earlier era of the web.

The problem isn't technical; it's organizational. When a Chief Marketing Officer (CMO) hears "AI agents," they think about whether their brand gets cited when customers ask Perplexity or ChatGPT a question. When a Chief Information Officer (CIO) hears the same phrase, they think about internal productivity tools and security threats. Both are talking about the same technology, but solving for completely different business problems, and that gap is now a revenue issue.

What Are the Three Types of AI Agents Hitting Your Website Right Now?

Understanding the distinction between different AI agent types is critical for both marketing and IT teams. Each layer serves a different purpose and requires a different strategy.

• AI Training Crawlers: Bots like GPTBot and ClaudeBot that pull content in real time for live conversations with users, not for later indexing. These need fast, structured, machine-readable pages to function effectively.
• AI Search Agents: Tools like Perplexity Comet and OpenAI Atlas that see pages on a user's behalf, compare products, fill forms, and initiate purchases. If your pages aren't machine-readable, these agents may skip your site entirely.
• AI Assistant Interfaces: ChatGPT, Claude, and Gemini, where humans still ask questions and read answers, but the line between "assistant" and "actor" is blurring rapidly as these tools take more autonomous actions.

Of these three layers, the first one is hitting your website right now and shaping how consumers discover and buy from you. Yet most CMOs and CIOs are mislabeling it as a security threat rather than a revenue opportunity.

Which AI Agents Are Actually Dominating Your Traffic?

The crawler data reveals a clear hierarchy of AI agents, with OpenAI's tools dominating across nearly every category. ChatGPT's user agent now accounts for more than 96% of AI user bot traffic, hitting sites in real time on behalf of consumers. GPTBot represents roughly 55% of AI training crawl volume, while OAI-SearchBot accounts for about 47% of AI search crawl activity.

But OpenAI isn't the only player growing. Applebot accounts for about 30% of AI search crawl activity, yet most search and marketing teams aren't even tracking it. ByteSpider, the crawler behind ByteDance's AI products, grew 138% over the tracking window. Most striking: ClaudeBot surged 800% between November and December 2025, signaling that AI training doesn't follow a linear schedule and that log analysis after the fact is no longer sufficient.

Google NotebookLM grew 144% in the same period as researchers and knowledge workers increasingly use it to pull live content on their behalf. And now Gemini user agents are surging as well. Of the 20% of brands with any agent policy set, 77% only block training crawlers. That's a publisher strategy: protect content from being learned by an AI language model. For a brand, the trade-off is different. Block training crawlers and the models never learn your story, handing the narrative to a competitor instead.

How to Align Your Marketing and IT Teams on AI Agent Strategy

Managing AI agent access has moved well beyond the SEO function. It now belongs jointly to marketing, IT, and the digital organization, and the brands moving fastest are treating it that way.

• CMO Responsibility: Gain clarity on how agents are shaping discovery and the brand impression customers form. Track how each AI engine talks about your brand, since these summaries often become the first impression a buyer ever forms of you.
• CIO Responsibility: Review bot policy and access controls through a revenue lens, not just a security one. If your robots.txt or firewall is keeping modern AI agents out, you're not screening bots; you're turning away customers.
• SEO and Digital Leadership: Ensure the pages that drive consideration and conversion are findable, machine-readable, and current enough for a machine to act on. None of these three roles can deliver the outcome alone.

The same AI engines sending agents to your site are also forming opinions about your brand and serving them back to customers as answers. AI engines now act as editorialists, each one summarizing your brand a little differently. Tracking how each engine talks about your brand has stopped being a research curiosity; it's now a revenue safeguard.

Why Are Most Companies Still Unprepared?

A survey of just over 1,000 enterprise digital and search marketing leaders reveals a consistent gap: awareness is broad, but ownership is unclear, and almost nobody can prove they're ready.

Only 19% could confidently answer "yes, we are ready, and I can prove it" if asked directly. Half are still working on it or unable to explain the gap to leadership. The ownership problem is even more striking: 75% have no documented plan or named owner for AI agent strategy. And 72% reported that marketing has ended up owning AI agent and Answer Engine Optimization (AEO) responsibility without ever being formally handed it. Only 17% report that IT or engineering owns the question.

Just 21% have built any strategy for search-side crawlers like OAI-SearchBot, and only 38% have any approach to user-facing agents browsing a site live on a customer's behalf. Most companies are fortifying the surface that matters least for revenue and leaving the two that matter most without a strategy at all.

What's the Real Cost of Getting This Wrong?

The financial stakes are substantial. Even if 80% of companies manage their AI agent policies correctly, the remaining 20% still leaves an estimated $40 billion of search opportunity on the table across the wider economy.

Brands have spent two decades engineering websites for human visitors and now must design for two audiences in parallel: humans and the AI agents acting on their behalf. The implication for IT leadership is uncomfortable but worth stating clearly: if your security policies are keeping modern AI agents out, you're not screening bots. You're turning away customers and making it harder for the marketing team to hit brand-building and revenue targets. And that impacts every department and function within your organization.

What Does Perplexity's IPO Timeline Tell Us About the Industry?

While the broader AI industry moves toward major public offerings, Perplexity is maintaining its own path. CEO Aravind Srinivas stated that the company is sticking with its 2028 IPO target, regardless of how OpenAI and Anthropic perform in their upcoming debuts.

"Agnostic of these two companies, we were planning for something in 2028, so that still remains the case," Srinivas told CNBC in an interview that aired on Tuesday.

Aravind Srinivas, CEO of Perplexity

Srinivas acknowledged that the performance of OpenAI and Anthropic's IPOs will matter to the broader industry. "I certainly think there will be ripple effects if they don't go well, like there is no sugar coating on that," he said. However, he defended the lofty valuations attached to both companies, arguing that they remain at the cutting edge of AI development.

Srinivas

At the same time, Srinivas suggested that investors will continue to scrutinize whether leading model developers can maintain a rapid pace of improvement. "If for six months you don't see a model capability advance from one of these two companies, then it's a problem for them," he noted. The debate over AI valuations comes as companies take a harder look at their AI budgets, with customers becoming more selective about which models they use.

Srinivas