D3sign | Moment | Getty Images
Bots have taken over the internet.
The State of AI Traffic report released Thursday by Human Security, a cybersecurity firm, showed that artificial intelligence and bots have officially eclipsed human users.
“The internet as a whole was created with this very basic notion that there’s a human being on the other side of the computer screen, and that notion is very rapidly being replaced,” Stu Solomon, CEO of Human Security, told CNBC.
That automated traffic across the internet grew almost eight times faster than human activity in 2025, according to the report.
Automated traffic, defined by Human Security as “internet traffic generated by software systems (including AI) rather than human users,” has surged rapidly as people continue to turn to AI chatbots for daily questions.
Proliferation of large language models like OpenAI’s ChatGPT, Anthropic’s Claude and Google’s Gemini has played a large part in AI traffic, which increased 187% from January to December 2025 according to the report.
“Machine-based traffic is effectively replacing humans as the dominant form of traffic on the other side of the internet,” Solomon said.
Human’s report was based data from its Human Defense Platform product, which it says processed over one quadrillion interactions across its customers. The process of quantifying automated activity across the entire internet can pose challenges, however, since there isn’t one complete database of interactions.
“You can try to estimate the amount of bot traffic by looking at the agent strings, but these are very noisy estimates,” Filippo Menczer, a professor of Informatics and Computer Science at Indiana University, told CNBC.
“They depend on what sample you get. They are depending on where you’re getting the data, where the measurements are coming from,” Menczer said.
User-agent strings are self-identified labels from web crawlers.
Human’s report acknowledges that while it used user-agent strings to identify AI operators, “the reliability of that self-identification is a growing concern.”
The report also addresses the surge in agentic activity from AI agents like OpenClaw that perform actions autonomously for a user. While 2024 agentic volume was extremely low, Human Security saw that traffic grow almost 8,000% in 2025.
Automated traffic includes popular features like Google‘s AI Overview and autofill, and is not necessarily malicious.
“This notion of machine bad, human good just is not realistic,” Solomon said. “You have to live in a world where machines are acting on our behalf, and we have to establish a level of trust that’s persistent over time.”
Though the Human Security report isn’t comprehensive, it serves as a significant benchmark in the AI age of the internet.
The industry has tracked the steady climb of automated traffic since the launch of ChatGPT in 2022.
At the SXSW conference last week in Austin, Cloudflare CEO Matthew Prince said that the internet was about 20% bot traffic prior to the generative AI era, and was mostly driven by Google’s web crawler.
Prince predicted that AI bots would exceed human traffic by 2027, citing “the rise of generative AI and its just insatiable need for data.”
