AI tools are being increasingly abused to launch cyberattacks

2 weeks ago

Computer generated look amongst absurd lines and points signifying AI

(Image credit: Shutterstock / IrenaR)

A increasing number of cyberattacks are being launched pinch nan thief of Artificial Intelligence (AI) and Large Language Models (LLM), caller investigation has claimed.

A study from Imperva noted betwixt April and September 2024, its Threat Research squad analyzed thousands of attacks, uncovering unit sites collectively acquisition much than 500,000 AI-powered attacks each day.

These attacks, nan researchers explain, often originate from AI tools specified arsenic ChatGPT, aliases Gemini, alongside bots designed to scrape websites for LLM training data. Cybercriminals were said to beryllium utilizing these devices mostly successful business logic maltreatment attacks, DDoS attacks, bad bots attacks, and API violations.

Business logic attacks

Business logic maltreatment was described arsenic nan astir communal AI-driven attack, taking up almost a 3rd (30.7%) of each incidents. It involves abusing morganatic features of different apps and APIs to transportation retired cyberattacks. DDoS are a adjacent 2nd (30.6%), while bad bot attacks return up a 5th (20.8%). The bots are designed to scrape pricing data, tally credential stuffing, arsenic good arsenic inventory hoarding.

"In erstwhile years, we've seen information threats for illustration Grinch bots and DDoS attacks origin awesome disruptions during nan vacation shopping season, affecting some retailers and consumers alike. Now, pinch nan wide readiness of generative AI devices and LLMs, retailers are contending pinch a caller activity of blase cyberthreats,” said Nanhi Singh, General Manager of Application Security astatine Imperva.

Singh added unit businesses request robust defenses, and a broad strategy, otherwise, they are risking losing delicate individual information, including in installments paper details, people’s addresses, and different relationship information. Identity theft and akin attacks tin lead to a tarnished image, nonaccomplishment of business, lawsuits, and regulator fines.

More from TechRadar Pro

The emergence of identity-related cyberattacks: costs, challenges and nan domiciled of AI
Here's a database of nan best firewalls today
These are nan best endpoint protection tools correct now

Sign up to nan TechRadar Pro newsletter to get each nan apical news, opinion, features and guidance your business needs to succeed!

Sead is simply a seasoned freelance journalist based successful Sarajevo, Bosnia and Herzegovina. He writes astir IT (cloud, IoT, 5G, VPN) and cybersecurity (ransomware, information breaches, laws and regulations). In his career, spanning much than a decade, he’s written for galore media outlets, including Al Jazeera Balkans. He’s besides held respective modules connected contented penning for Represent Communications.

Source Technology