Agentic AI Security: Governments Demand Stronger Controls

Government agencies are urgently calling for stronger security around Agentic AI. The Australian Signals Directorate just released new guidance to address significant cybersecurity risks. Here's the thing: Agentic AI systems are taking on bigger roles in businesses. Gartner predicts 33% of enterprise software will use them by 2028, up from less than 1% today. But there's a big problem. A staggering 88% of Australian organizations expect AI agents to outpace their security safeguards this year. What's interesting is that while AI offers huge productivity gains, it also creates new vulnerabilities. Microsoft's report shows these AI agents need strong protection, just like human users. Rubrik found that over 80% of companies say AI agents demand more manual oversight than the efficiency they deliver. The fear is that these systems will drive most cyberattacks in the coming year. The bottom line: As decision-making shifts from people to machines, keeping things safe and secure is a critical challenge for everyone.