Security Expert Warns: AI Agents Could Be Your Biggest Security Risk

A top security expert is sounding the alarm about AI agents — computer programs that can act on their own to complete tasks for you.

Ronghui Gu, CEO of CertiK (a company that checks blockchain security), warns that companies and people are rushing to use AI agents without proper safety measures. These AI agents are like digital assistants that can access your files, passwords, and even your bank accounts to help you work faster.

The problem? These AI agents could be easily hacked or tricked. When you give an AI agent access to your computer or accounts, it's like giving a stranger the keys to your house. If someone finds a way to control that AI agent, they could: • Read your private files • Steal your passwords • Access your money accounts • Use your identity to commit fraud

Gu explains that many people trust these AI agents too much. "Right now, agents are no longer just answering questions in a chat window," he said. "They are beginning to call external tools, read local files, trigger workflows, and interact with financial infrastructure."

The security risks are already showing up. CertiK's research found widespread vulnerabilities (weak spots that hackers can exploit) and a surge in automated scams targeting other AI systems. Some experts predict that by 2035, there will be more AI agents than humans on the internet.

The solution? Gu recommends treating AI agents like you would any unknown program — scan them for viruses, test them in isolated environments (separate from your main systems), and never give them access to sensitive information until you're sure they're safe.

This is an AI-generated summary. Read the original article at: https://www.coindesk.com/tech/2026/05/29/mass-deployment-of-ai-agents-is-a-disaster-waiting-to-happen-says-certik-ceo