AI Agents Collapse Bot/Human Binary, Driving 7,851% Traffic Surge
HUMAN Security's 2026 report reveals that AI-driven traffic grew 187% in 2025, with agentic AI traffic increasing by 7,851%. This shift has collapsed the traditional bot/human binary, complicating fraud detection and revenue capture by making benign automation mechanically indistinguishable from malicious automation. Streaming and media is identified as one of the top three verticals affected, with implications for fraud (account takeover, carding, scraping, fake accounts) and ad tech.
Key Takeaways
- AI-driven traffic grew 187% in 2025; agentic AI traffic, which transacts autonomously, increased 7,851%.
- Streaming and media, retail, and travel verticals accounted for over 95% of AI-driven traffic growth.
- OpenAI bots (ChatGPT User, OAI-SearchBot, GPTBot, ChatGPT Agent) comprised 69% of observed AI-driven traffic.
- Agentic AI now executes transactions, with 2.31% of agentic activity on checkout pages and 8.82% on account pages.
- Web scraping volume rose 47% in 2025, with heavily targeted companies seeing over 61% of traffic as scraping attempts.
Why It Matters
The traditional bot-or-not security model is obsolete, forcing streaming platforms to distinguish trustworthy AI interactions from malicious ones without clear behavioral signals. This directly impacts revenue, fraud exposure, and content monetization strategies as AI agents increasingly browse, log in, and transact. With AI traffic growing eight times faster than human traffic and concentrated within a few major operators, access policy decisions about these companies now dictate the majority of an organization's AI exposure. Teams must move beyond blocking all automation to a nuanced understanding of intent, adopting tools that provide real-time visibility into AI agent activity post-login.
Read full article at ppc.land