This summer, Russian hackers executed a new phishing campaign targeting Ukrainians by embedding an artificial intelligence (AI) program in emails. Once installed, this malicious software would seek out sensitive files to relay back to Russia, marking the first use of large language models (LLMs) in cyberattacks. Hackers, criminals, and cybersecurity professionals have increasingly integrated AI tools into their operations. While LLMs like ChatGPT enhance efficiency, they have not fully transformed novice hackers into experts yet. Major cybersecurity firms, including Google, are actively utilizing AI to identify vulnerabilities more swiftly. However, the advantage currently seems to favor defenders, as AI tools help uncover overlooked bugs before they are exploited. There’s concern that once advanced AI hacking tools are freely available, the landscape could shift drastically. The rise of agentic AI—the capability for more complex tasks—poses a potential insider threat by enabling actions without sufficient safeguards. Cybersecurity remains a dynamic cat-and-mouse game as technology evolves.
Source link
Share
Read more