Artificial intelligence is changing the cybersecurity landscape in the United States, as increasingly capable AI tools help researchers uncover software vulnerabilities while potentially giving cybercriminals new ways to attack digital systems.

Recent incidents have highlighted the growing challenge. A critical vulnerability affecting the widely used PHP programming language prompted emergency security measures, including rapid software updates and temporary service disruptions at some election-related technology providers.

Security researchers are increasingly using advanced AI systems to identify weaknesses that traditional testing methods might miss. However, experts warn that attackers could also use similar capabilities to discover vulnerabilities, automate attacks and target organisations more efficiently.

The development has attracted particular attention ahead of the US midterm elections, with election officials reviewing the security of websites, registration services and other supporting infrastructure.

Technology companies are responding with stronger vulnerability monitoring, security testing and cooperation with government agencies. The central challenge is ensuring that cybersecurity defenders can use advanced AI capabilities before malicious actors exploit them.

As businesses, government agencies and critical infrastructure providers become more dependent on digital systems, AI-driven cybersecurity is emerging as an increasingly important national technology priority.