Will the step forward in frontier AI mean a step backward in cybersecurity?

ON APR 15, Singapore’s Cyber Security Agency (CSA) issued an advisory warning that frontier artificial intelligence models can reduce the time required to identify vulnerabilities and engineer exploits – from months to hours.

The emergence of these models is a cause for concern for all organisations. The potential to identify vulnerabilities in code will result in increased volume and speed of attacks on IT infrastructure.

According to a recent VulnCheck study, 28.73 per cent of zero-day exploits – vulnerabilities unknown to developers and defenders – happen within 24 hours. Frontier AI models could potentially worsen the situation for defenders.