Artificial Intelligence | News, how-tos, features, reviews, and videos
By exploiting how AI coding agents retrieve and verify plugins, researchers were able to execute malicious code even when the agent was told to use a trusted, approved version.
Agents’ misbehaviors included prompt injection, covert communication, and credential searches.
AI is becoming the biggest destination for new security dollars, but spending remains uneven and strongest among organizations with mature AI governance.
Hosting open-weight AI models locally can offer enterprises more control — unless an AI agent decides to change the model that powers it.
Diverging safety approaches could fragment access and complicate enterprise AI strategy.
Stronger monitoring, shorter-lived tokens, and tighter controls over how tokens are used after authentication are needed as AI agents complicate who and what enterprises can trust inside their systems.
Agent-run SOCs, machine-speed containment, and flatter security teams could transform how cyber defense is practiced as the AI era takes hold.
Analyst argues that OpenAI’s depiction of the agents as 'benign' may pose a greater threat than the attack itself, generating SOC AI alert fatigue.
The new offering correlates AI-agent activity with endpoint, identity, cloud, SaaS and code data to help security teams detect and respond to threats.
