Enterprises face the highest risk from cyber assaults executed using artificial intelligence for the third quarter in a row, as revealed in research by Gartner.
A study conducted by the consulting agency from July to September, involving 286 senior executives in risk and assurance roles, indicated that 80% identified AI-driven malicious attacks as their top concern. The surge in AI-assisted attacks comes as no surprise, given the increasing evidence of their prevalence.
Additional emerging risks highlighted in the report include AI-driven misinformation, rising political divisions, and discrepancies in organizational talent profiles.
AI is being exploited by cyber criminals to develop malware and phishing emails, among other activities
In a recent incident, HP intercepted a malware-disseminating email campaign utilizing a script thought to have been authored with the aid of GenAI. The VBScript displayed a well-organized structure, complete with comments for each command, a level of detail that would be excessively meticulous for a human to undertake.
By replicating the process with GenAI, researchers obtained a similar script output, indicating that the original malware was potentially partly AI-generated.
SEE: 20% of Generative AI ‘Jailbreak’ Attacks are Successful
During the second quarter, security company Vipre detected a 20% increase in business email compromise attacks compared to the same period in 2023, with two-fifths of these attacks being AI-generated. Chief Product and Technology Officer of VIPRE, Usman Choudhary, stated in a press release that malefactors are now leveraging advanced AI algorithms to fabricate convincing phishing emails, capturing the tone and style of legitimate communications.
According to Imperva Threat Research, retail platforms encountered an average of 569,884 AI-driven attacks each day between April and September. Researchers revealed that tools like ChatGPT, Claude, and Gemini, as well as specialized bots that scrape websites for LLM training data, are being utilized to carry out distributed denial-of-service assaults and business logic manipulation.
An increasing number of ethical hackers are acknowledging their utilization of GenAI, with the proportion rising from 64% to 77% over the past year, according to a report from BugCrowd. These researchers suggest that AI aids in die-channel attacks, fault-injection attacks, and automates parallelized assaults to breach multiple devices simultaneously, hinting at the significant value AI provides, not just to defenders but to malicious actors as well.
The increase in these threats should not come as a shock
AI facilitates a reduction in the barrier to entry for cybercriminal activities, enabling less skilled perpetrators to utilize it for various purposes such as generating deepfakes, network scanning for vulnerabilities, conducting reconnaissance, etc. Researchers at ETH Zurich recently developed a model capable of solving Google reCAPTCHAv2 puzzles, used to differentiate between humans and bots, with 100% effectiveness.
Analysts at the security firm Radware predicted at the beginning of the year that this enhanced accessibility would result in the development of private GPT models for illicit objectives. They also projected a rise in the number of zero-day exploits and deepfake scams as malicious actors become more adept with LLMs and generative adversarial networks.
Google’s Mandiant recorded a total of 97 zero-day vulnerabilities identified and exploited in 2023, marking a 56% surge from the previous year. Recently, Microsoft included deepfakes among the most significant attack vectors exploited by increasingly prolific ransomware groups.
SEE: AI Deepfakes Rising as Risk for APAC Organisations
Executives also express concern about relying too much on IT service providers
For the first time, the criticality of IT vendors surfaced as a top concern among senior risk and assurance executives, according to Gartner.
Zachary Ginsburg, Senior Director of Research within the Gartner Risk and Audit Practice, highlighted in a Gartner press release: “Organizations with a heavy reliance on a single vendor may encounter elevated risk in case of service disruptions or may face unexpected service alterations due to new regulations or legal decisions in regions like the EU or U.S.”
He referenced the CrowdStrike outage in July, which resulted in approximately 8.5 million Windows devices globally being incapacitated, causing significant disruptions to emergency services, airports, law enforcement agencies, and other essential entities.
SEE: What is CrowdStrike? Everything You Need to Know
About Author
