The UK’s National Cyber Security Centre (NCSC) has issued a report highlighting the potential for AI to significantly escalate cyber threats within the next couple of years.
The report specifically points to an expected increase in ransomware attacks, where attackers use malware to encrypt the files or systems of their victims and demand a ransom for the decryption key.
According to the NCSC, AI is set to boost the effectiveness of cybercriminals, particularly by enabling more convincing phishing schemes that deceive people into handing over sensitive data or engaging with harmful links.
The report notes, “Generative AI can already create convincing interactions like documents that fool people, free of the translation and grammatical errors common in phishing emails.”
It identifies the progression of generative AI, which can produce realistic interactions and documents without the typical signs of phishing attempts, as a major factor in the heightened cyber threat level anticipated over the next two years.
The assessment also points to the increasing challenges in maintaining cyber resilience, such as the complexities in authenticating emails and password reset inquiries. This is compounded by the decreasing interval between the release of security patches and the exploitation of vulnerabilities, making it tougher for network administrators to apply fixes promptly.
James Babbage, the National Crime Agency’s director general for threats, stated: “AI services lower barriers to entry, increasing the number of cyber criminals, and will boost their capability by improving the scale, speed, and effectiveness of existing attack methods.”
Despite these concerns, the NCSC report suggests that AI could also enhance cybersecurity measures by better detecting attacks and improving system architecture. It advocates for continued exploration into how AI-driven defensive technologies can counter these growing risks.
The report acknowledges that, at present, the sophistication required for AI-enhanced cyber operations is mostly within the reach of highly skilled state entities. However, it cautions that these barriers are diminishing as such capabilities become commercialized and accessible to a broader range of malicious actors.
