D3 Sign | Moment | Getty Images
Bots have taken over the internet.
The State of AI Traffic report released Thursday by cybersecurity firm Human Security showed that artificial intelligence and bots have officially overtaken human users.
“The whole Internet was built on a very basic concept of there being a human on the other side of the computer screen, and that concept is being replaced very quickly,” Stu Solomon, CEO of Human Security, told CNBC.
Automated traffic, defined by Human Security as “Internet traffic generated by software systems (including AI) rather than human users,” is rapidly proliferating as people continue to rely on AI chatbots to ask everyday questions.
According to the report, automated traffic on the internet grew nearly eight times faster than human activity in 2025.
The proliferation of large-scale language models such as OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini are playing a major role in AI traffic, which increased by 187% from January to December 2025, according to the report.
“On the other side of the Internet, machine-based traffic is replacing humans as the dominant form of traffic,” Solomon said.
Human’s report is based on data from its Human Defense Platform product, which the company says has processed more than 1,000 trillion interactions across its customers. However, the process of quantifying automated activity across the Internet can be challenging, as there is no single complete database of interactions.
“You can estimate the amount of bot traffic by looking at the agent strings, but this is a very noisy estimate,” Filippo Mentzer, a professor of informatics and computer science at Indiana University, told CNBC.
“They depend on what kind of samples you get. They depend on where you get the data, where the measurements come from,” Mentzer said.
A user agent string is a self-identifying label from web crawlers.
Human’s report acknowledged that it had been using user agent strings to identify AI operators, but that “the reliability of that self-identification is increasingly concerning.”
The report also highlights the surge in agent activity from AI agents like OpenClaw that autonomously perform actions on behalf of users. While agent volume was very low in 2024, Human Security predicted that traffic would increase by almost 8,000% in 2025.
Automated Transportation includes popular features such as: googleAI summary and autofill are not necessarily malicious.
“This concept of machines being bad and humans being good is completely unrealistic,” Solomon said. “We have to live in a world where machines act on our behalf, and we have to establish a level of trust that will last over time.”
Although the Human Security Report is not comprehensive, it serves as an important benchmark in the Internet’s AI era.
The industry has been tracking a steady increase in automated traffic since the launch of ChatGPT in 2022.
At the SXSW conference held in Austin last week, cloudflare CEO Matthew Prince said that before the era of generative AI, bot traffic on the internet was about 20%, and most of it was driven by Google’s web crawler.
Prince predicted that AI bots will overtake human traffic by 2027, citing “the rise of generative AI and its insatiable demand for data.”
