More than half of internet traffic is already automated. Now layer in AI agents that are writing RFPs, fact-checking your marketing, and starting to make purchase decisions on behalf of real humans. The bot conversation has changed.
The question stops being how do we block bots? It becomes which automated traffic is good for the business, which is bad, and who owns the call?
That’s the question at the center of recent Forrester research, and it’s the same question we’ve been pushing security and marketing leaders to answer together for a while now.
The report is Determine Trustworthy Business Traffic From AI Agents, Bots, And Scrapers To Get Marketing And Security Right (Forrester Research, Inc., February 24, 2026). The message is direct. Marketing and security can’t keep operating in parallel on automated traffic. They have to operate together.
The cost of staying in silos is now a P&L item
According to Forrester, when marketing and security don’t collaborate around AI agents, bots, and scrapers, the business takes four concrete hits.
Ad spend gets wasted on impressions and clicks that will never convert.
The funnel gets distorted, because automated traffic gets miscounted as either too much demand or too little, and revenue forecasts follow.
Premium content gets scraped without consent, eroding the monetization model the business just spent a decade building.
And legitimate buyer-assist agents get blanket-blocked, killing deals before sales ever sees them.
Security can’t fix any of these alone. Neither can marketing. Each problem sits between the two functions, and each one costs real money.
“Block everything” and “let it all through” are both wrong
Security looks at automated traffic and sees risk, so the instinct is to block. Marketing looks at the same traffic and sees engagement, so the instinct is to count it as performance. Both instincts fail with AI agents in the mix.
The same automation can be a legitimate demand signal or a hostile scraper depending on intent. A buyer-assist agent helping a procurement team build an RFP shortlist is exactly the kind of session you want on the site. An LLM scraper pulling your premium content into someone else’s training run is the opposite.
You can’t tell those apart with marketing analytics. You can’t tell them apart with security tooling either, not without business context that only marketing can supply.
Forrester’s blueprint, and why we agree with it
The most useful thing in the report is that Forrester doesn’t stop at “you should collaborate.” It lays out a five-phase blueprint: shared taxonomy, detection and classification, measurement and attribution, response and orchestration, and governance and collaboration through a cross-functional working group.
Detection alone creates friction. Strategy without detection leaves exposures you can’t see. The fix is shared infrastructure that both teams trust. A single read on which requests are human, which are helpful automation, and which are hostile, before any of it hits a campaign report or a fraud queue.
That’s the foundation we’ve been building toward for years, and it’s exactly what AI Agent Trust is designed to do. Verify agent identity and intent in real time, so security and marketing can act on the same signal.
What this means for security and marketing leaders right now
If you lead security, your detection is only as valuable as the business decisions it informs. Your marketing counterparts need agent share, intent patterns, and fraud events feeding their analytics in a language they can act on.
If you lead marketing, your campaign metrics are only as honest as the traffic underneath them. You need a security partner who can tell you which sessions are human, which are helper agents, and which are noise. Before you optimize against the wrong signal, not after.
Another tool won’t solve this. A shared playbook will. Forrester has the playbook. Kasada has the runtime.
Read the full Forrester report →
Source / Attribution
Determine Trustworthy Business Traffic From AI Agents, Bots, And Scrapers To Get Marketing And Security Right, Forrester Research, Inc., February 24, 2026.
Information in Forrester publications is based on Forrester’s efforts to compile and analyze the best resources reasonably available to Forrester at any given time. Opinions reflect judgment at the time and are subject to change. This report is part of a broader collection of Forrester resources, including interactive models, frameworks, tools, data, and access to analyst guidance.
