AI-driven bot activity surged in 2025, increasing roughly 12.5x year over year, with AI agents now a discrete traffic category alongside traditional good and bad bots.

Financial services bear the brunt of automation-enabled crime, accounting for nearly a quarter of bot attacks and nearly half of account takeover incidents.

Singapore-specific data show financial services accounting for 79% of bot attacks, with computing and IT sectors leading account takeover incidents at 45%.

Thales’ executive notes the challenge has shifted from merely identifying bots to understanding bot intent and its alignment with business goals and critical system interactions.

Attack patterns are increasingly legitimate-seeming, often using valid credentials and well-formed requests while abusing business logic to extract data or manipulate workflows at scale.

AI-driven automation is reshaping the digital landscape; traditional bot defenses focused on detection and blocking are insufficient in automated environments where governance, visibility, policy enforcement, and behavioral analysis are needed.

APIs and identity systems are primary targets, with about 27% of bot attacks aimed at APIs to bypass front-end defenses and exploit backend logic at machine speed.

AI agents interact with applications and APIs to fetch data, complete tasks, and automate workflows, creating a classification challenge since automated activity can be legitimate, harmful, or hard to verify from traffic patterns.

The rise of AI-driven activity creates a visibility gap for organizations, complicating security risk assessment and real-time threat detection.

Global internet traffic now includes more than half from AI-driven automation, with roughly 40% of that traffic identified as malicious, underscoring a shift toward intent-driven security challenges.

As bots become active participants in digital ecosystems, effective management of automation is essential to maintain security and trust in online services.

Bot traffic exceeded 53% of web traffic in 2025, up from 51% the previous year, indicating bots outnumber humans and operate more continuously.