AI-Driven Vulnerabilities Threaten Global Internet Security, Experts Warn
Artificial intelligence models are helping drive a doubling of discovered software vulnerabilities, creating a wave of code flaws that cybersecurity professionals warn will trigger disruption, according to a Reuters report. The surge follows a week of major disruptions across government websites in Europe and elsewhere.
The recent digital disruptions began on Wednesday after operators discovered severe weaknesses in PHP, a foundational programming language powering a large portion of the internet. Faced with potential mass exploitation, administrators scrambled to patch systems while others pulled their sites offline entirely. VulnCheck researcher Patrick Garrity noted that the PHP scramble coincided with disclosures in Node.js, Plex, LiteLLM, Railway, and an eMudhra cryptographic key issue.
Matthew Prince Warns of Weekly Crises
Matthew Prince, chief executive of digital infrastructure firm Cloudflare, stated in an interview that he is tracking numerous additional severe vulnerabilities in critical software. Prince predicted that AI-driven discoveries of long-hidden internet weaknesses will make every week feel like a crisis for the next couple of years.
I think we’re going to have a couple years where every single week it’s going to feel like the sky is falling.
— Matthew Prince, Chief Executive, Cloudflare
According to VulnCheck research slated for publication in coming days, automated threat intelligence shows AI tools have effectively doubled detected digital flaws compared to previous periods. Shane Fry, chief technology officer at RunSafe Security, emphasized that AI models have altered industry timelines by conducting round-the-clock code scrutiny and generating working exploits within hours rather than months.
Experts Divided on Hacker Success and Weaponization
Despite faster vulnerability discovery and exploit weaponization, security experts remain divided on whether the trend will cause widespread successful breaches. Daniel Stenberg, lead developer of the file transfer tool curl, argued that AI-discovered flaws often involve exotic bugs that are difficult for rogue actors to exploit.
ArmorCode and Hunt.io Assess Immediate Threat Levels
Ramy Rahman, senior principal solutions engineer at ArmorCode, supported this view by noting that increased disclosures do not automatically translate to successful attacks. Threat intelligence firm Hunt.io reported aggressive exploitation attempts against the recent PHP vulnerability by Amazon, Cloudflare, and others, but confirmed that cybercriminals had failed to assemble a working exploit tool by Thursday afternoon.
Cloudflare chief Matthew Prince reported seeing anecdotal evidence of minor compromises tied to the PHP bug. However, he maintained that surface-level exposure of these flaws acts as a necessary cleansing process for global digital infrastructure.