GenAI Both An Enabler And Destructor; Companies Wary Of AI-Fuelled Malware, Phishing: Report

GenAI presents dual prospects of possibilities and peril, and a large majority of cybersecurity leaders from around the world are wary of its negative impact, a global survey has found.

A staggering 93 per cent of respondents to the survey conducted by CyberArk said they anticipate a negative impact from AI, with AI-powered malware and phishing topping the list.

CyberArk's 2024 Identity Security Threat Landscape Report -- which surveyed 2,400 cybersecurity leaders across 18 countries -- found that "cyber debt continues to build with GenAI the rise of machine identities, and increasing third and fourth-party risks".

The report said about 99 per cent of organisations take the help of AI in cybersecurity defence initiatives.

However, it also predicted "an increase in the volume and sophistication of identity-related attacks, as skilled and unskilled bad actors also increase their capabilities, including AI-powered malware and phishing".

A total of 93 per cent of respondents, it said, anticipate AI-powered tools to create cyber risk for their organisations in FY25.

"In the last 12 months, 9 of 10 organisations were victims of a breach due to a phishing/vishing attack. These types of attacks will be harder to detect as AI will automate and personalize the attack process," it said.

Altogether, 93 per cent of Indian organisations had two or more identity-related breaches in 2023, it said.

"Looking at the year ahead, organisations can expect to be affected by data leakage from compromised AI models, AI-powered malware, and phishing," it added.

The report further raised concerns about the burgeoning circulation of deepfakes, especially with elections around the corner.

"Get ready for a pivotal and unsettling Decision 2024 as over 4 billion voters prepare to elect leaders in over 60 countries -- and AI-powered deepfake campaigns become the weapon of choice for anyone wanting to influence election outcomes," it said.

The report also named machine identities as a vulnerable source of risk, "ready to be exploited by bad actors with AI-powered ability to execute at scale".