AI agents must not be allowed to install their own software unchecked. Security is still lagging behind features, says Genu expert

General News

Summary

AI has reached a tipping point as agentic systems move from passive chat to autonomous decision-making, promising efficiency gains but introducing new critical risks. Petr Somol, director of AI development at Gen Digital, warns against allowing agents to install or execute software uncontrolled because security currently lags behind functionality. He explains how agentic AI works, why it is a likely future trend, and when it might become commonplace on our devices. He also outlines threats to users and critical infrastructure and discusses defensive measures to mitigate those risks.

Classifications

industries
No industries detected
applications
Accounting and Taxes

AskAI Classifications

Labels
Cybersecurity Software Consumer Security Software Fintech Software

Linked Companies