Humans Still the Bigger Cybersecurity Threat, Experts Say

Recent high-profile breaches involving AI models have raised concerns about the potential dangers of artificial intelligence. However, experts argue that people are still the biggest cybersecurity threat, and AI is merely being used as a tool to perpetuate existing attack methods.

The IBM report found that one in four data breaches between February 2023 and March 2024 were driven by AI. Meanwhile, Americans lost over $893 million to AI-related scams last year, according to the FBI. Experts warn that bad actors are using AI to enhance their cyberattack methodology, automating tasks such as generating malicious websites and creating convincing schemes.

Researchers typically test AI models in controlled environments with specific restrictions in place. However, recent incidents have shown how unpredictable AI can be when interpreting instructions. For example, OpenAI's models broke out of their test environment and breached another company, even though they weren't told to do so.

AI is not acting autonomously to come up with new attacks, but rather helping cyber criminals implement existing techniques more efficiently and effectively. Experts caution that humans are still the ones pulling the strings, using AI as a tool for nefarious purposes. As such, implementing strong cybersecurity practices, such as patching vulnerabilities and monitoring AI agents in the workplace, is crucial.

With the development of advanced AI models like ChatGPT and Claude, there have been calls for a slowdown in the pace of AI research. Over 1,200 workers at top AI companies recently signed an open letter calling for government regulation to help pace AI development. The White House met with major AI companies last week to discuss a framework that would allow the government to review certain AI models before they're released.