How to stay safe when AI is breaking the Internet

Picture Credit: by CNN

The discovery of long-standing vulnerabilities in PHP, a programming language that powers around 70% of the world’s websites, has renewed concerns about the security of the internet’s underlying infrastructure. Although the flaws had existed for more than a decade, researchers were able to better understand how attackers could exploit them with the help of artificial intelligence.

Known as PHP2Shell, the vulnerability could allow hackers to remotely execute code or access information stored on affected servers. Its discovery prompted organisations to urgently update their systems, with some government websites in Europe temporarily going offline while security fixes were installed.

The incident highlights a growing challenge for cybersecurity teams: much of the world’s digital infrastructure relies on ageing technologies that may contain undiscovered weaknesses. As AI makes it easier for researchers and malicious actors to identify vulnerabilities, similar security incidents could become more common.

While individuals cannot directly protect websites from server-side flaws, experts stress that basic cybersecurity practices remain essential.

Installing software updates promptly is one of the simplest ways to reduce exposure to known vulnerabilities. Users should also check whether their devices continue to receive security updates, particularly if they own older smartphones or computers.

Phishing attacks are another major concern. AI can help cybercriminals produce convincing messages with fewer spelling and grammatical errors, making fraudulent emails and texts harder to identify. Rather than clicking suspicious links, users should access accounts directly through official websites or contact the relevant organisation to verify unexpected requests.

Picture Credit: by Forbes

Strong password practices are equally important. Using unique passwords for every account, preferably generated and stored through a reputable password manager, can reduce the damage caused by compromised credentials. Where available, passkeys and two-factor authentication offer additional protection.

Users should also monitor important accounts for unusual activity and verify suspicious calls or messages through trusted contact channels.

Although AI is increasing the sophistication of cyberattacks, many still rely on familiar techniques designed to exploit human trust. Keeping software updated, recognising potential scams and strengthening account security remain practical steps individuals can take to protect themselves online.