Dell Technologies Warns of Security Risks from AI Agents!

As companies increasingly deploy AI agents to automate complex tasks within enterprise systems, cybersecurity experts are warning about the growing risks associated with these technologies. Speaking at the Dell Technologies World conference in Las Vegas, John Scimone, President and Chief Security Officer of Dell Technologies, said AI agents are highly effective and persistent in achieving assigned goals, but may create unintended security risks if not properly governed.

AI agents are capable of browsing the web, accessing files, calling APIs, and querying databases. However, experts highlighted that large language models still face vulnerabilities such as prompt injection attacks, where malicious instructions hidden within trusted data sources can manipulate AI systems. Researchers warned that poorly secured AI agents could potentially become a new form of insider threat within organisations.