New Age Cyber Threats Are Here—the US Must Be Prepared – The National Interest

New Age Cyber Threats Are Here—the US Must Be Prepared  The National Interest

As artificial intelligence rapidly expands the cyber threat frontier, public-private cooperation is needed to defend the digital domain. The US Capitol building at night, circa September 2025, with glowing digital padlock icons and data elements overlaid. US cybersecurity is under tremendous strain from new AI-enhanced threats, writes Pat Fallon. (Shutterstock/VideoFlow) Last week, several components of the US intelligence community and law enforcement apparatus, including the FBI Cyber Division and NSA Cyber team, issued notices that a major cyberattack, referred to as PHP2Shell, exploited a critical vulnerability that puts over 70 percent of websites at risk. Nearly every major website of every US industry is under threat, from software companies to large US banks to telecommunications providers and scientific labs.

These targets share the same programming language: PHP. This programming language is vulnerable because it has a low barrier to entry, making it easy for inexperienced developers to write code with common, easily exploitable flaws. Many companies have reported AI agents, also called agentic superintelligence (SI), attempting to breach their cyber defenses and disrupt customer operations; once inside, these agents can access sensitive private data. While details of last week’s attack are still emerging, two truths are clear: an attack of this scale puts nearly every American at risk, and the cybersecurity landscape has changed significantly in the age of agentic