Automated bots exceed half of internet traffic as AI attacks surge
Updated
Updated · letsdatascience.com · May 10
Automated bots exceed half of internet traffic as AI attacks surge
6 articles · Updated · letsdatascience.com · May 10
Imperva put bot traffic at 51% in 2025, while Thales estimated 53%, with 40% of bot traffic classified as malicious and Human Security reporting AI agent traffic up nearly 8,000%.
Thales said about 27% of bot attacks targeted APIs, while automated traffic grew roughly eight times faster than human traffic, increasing pressure on online services and fraud defences.
The reports suggest a broad shift toward humanlike, AI-assisted automation using tools such as headless browsers and proxies, though vendors caution internet-wide bot estimates vary by telemetry and methodology.
With bots outnumbering humans online, how can we ever trust digital interactions again?
As AI now writes its own cyberattacks, can our defensive AI possibly keep pace?
AI Bots Dominate 2026 Web: 40% of Traffic Now Malicious, Demanding New Security Models
Overview
The Thales 2026 Bad Bot Report reveals a dramatic shift in the digital landscape, with bad bots now making up 40% of all internet traffic. This surge in malicious automation is reshaping online interactions and marks a new era where automated traffic is a critical, often harmful, part of the digital ecosystem. Based on extensive data collected throughout 2025, the report highlights how advanced bots, powered by AI, are increasingly targeting vulnerabilities and challenging traditional security measures. As a result, organizations face growing threats that demand new strategies to protect digital infrastructure and maintain trust online.