Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI news ·

Advanced AI models pose new cybersecurity threat to unprotected organisations, expert warns

Advanced AI can complete a year's worth of penetration testing in under three weeks, and attackers will soon use the same tools at scale. Government agencies have roughly six months to update defenses before this threat class becomes widespread.

1 min read
Government

Advanced AI Models Will Force Government Agencies to Rethink Cybersecurity

Government agencies without updated security defenses will face a new class of threats within six months as advanced AI models become widely available, according to Lee Klarich, chief technology officer at Palo Alto Networks.

The cybersecurity firm tested emerging AI systems and found they excel at identifying vulnerabilities in code. In less than three weeks, the AI accomplished the equivalent of a full year's worth of penetration testing effort.

The capability gap matters because attackers will use these same tools. "Attackers will find the seams in those guardrails," Klarich said. "They will use advanced AI to discover zero-day vulnerabilities at scale, generate exploits in near real time, and develop autonomous attack agents unlike anything the industry has faced."

How Advanced AI Changes the Attack Surface

Advanced AI doesn't just find individual security flaws. It excels at "vulnerability chaining" - combining multiple lower-severity issues into critical-level exploits. One example: linking two medium-severity vulnerabilities and one low-severity flaw into a single critical attack path.

The AI can also analyze the full exposure surface of applications, including public-facing platforms, and identify vulnerabilities that traditional security tools miss.

What Government Organizations Need to Do Now

The standard approach to cybersecurity no longer works. "Organisations that are 'mostly protected' are effectively unprotected," Klarich said.

Government agencies should take these steps:

  • Use advanced AI models to assess your entire code and application landscape
  • Build a comprehensive inventory of assets and exposures
  • Prioritize finding and fixing vulnerabilities faster - legacy security operations timelines are no longer viable

Attack cycles are accelerating. Remediation and exposure reduction are now baseline requirements, not optional improvements.

For government cybersecurity professionals, the AI Learning Path for Cybersecurity Analysts provides structured training on defending against AI-driven threats and understanding how these systems work.

Share