Over half of internet traffic in 2023 was driven by artificial manipulators and content.

Artificial Accounts, Content and Fraud methods generated tens of billions of attacks in 2023 resulting in reputation damage, financial losses, operational disruption, market disruption and more. Malicious traffic accounting for almost 2/3 of all web traffic globally. Source: Arkose Labs

In a world where digital communication dominates, large public-facing companies and eminent figures grapple with an ever-escalating threat from sophisticated cybercriminal operations and organized criminal syndicates. 

They deploy advanced technologies capable of scraping data, creating fake accounts, and manipulating online discourse. 

These artificial tools are not only adept at evading traditional security measures but are also used to fine-tune Generative AI models for more effective phishing and social engineering attacks.

When automated efforts are stymied by security systems, these malevolent entities switch tactics, resorting to human fraud farms. Employing real individuals, sometimes in conditions bordering on coercion, these farms mimic authentic online behaviors, rendering detection exceedingly difficult. 

Their activities range from seizing control of accounts to creating bogus registrations and misusing products, posing severe risks to the integrity and reputations of high-profile organizations and individuals on social media.

Basic Agents

Limited artificial agents that perform basic tasks.

Intelligent Agents

Systems capable of complex, content aware interactions.

Human Fraud Farms

Organized networks of coerced or WFH labour in low-wage areas. They leverage solvers that use automation.

The Growing Impact of Generative AI

The proliferation of GenAI has transformed the cybersecurity landscape, lowering the barrier to entry for attackers. This technology is being used by bad actors for content generation, including creating sophisticated phishing emails. There’s an expectation of a major increase in romance scams on social media, as bad actors use GenAI to craft perfectly worded responses on dating apps and sites.

In the first half of 2023, a significant portion of traffic to digital properties, including social media, was malicious. This includes both intelligent agents, capable of sophisticated techniques like machine learning and AI to mimic human behavior, and basic tools.

Koat.ai Data

We observed a 121% increase in total attacks in Q2 over Q1 2023.

The artificial account and content problem is a weighty one that escalated exponentially in the first half of 2023—167% in Q2 2023. And when fraudsters’ systems are blocked, they often pivot the attack to human fraud farms. This new analysis shows that human-based attacks increased 26% in Q3 over the second quarter time period.