United States — A team of former Cisco artificial intelligence security researchers has launched Tenet Security Inc., a startup focused on stopping so‑called rogue AI agents from causing harm in enterprise environments. As more companies deploy autonomous AI systems that can execute code, query internal databases and make real-time business decisions, Tenet aims to address the rising risk that these agents will follow destructive or unintended objectives. The company says that handing critical infrastructure and sensitive data to autonomous systems without sufficient oversight has created a serious security gap for large organizations. United States — Tenet’s platform uses what it calls Agent-Side Simulation, a proprietary approach that tests an autonomous AI agent’s intended actions in a virtual environment before they run on live infrastructure. Instead of only alerting after suspicious behavior has already taken place, the system simulates the planned sequence of actions, scores the risk and, if it identifies a high-risk or malicious path, blocks the behavior and generates a trace explaining why it was flagged. The startup targets vulnerabilities such as unintended resource depletion, unauthorized data access and autonomous behavior drifting into unsafe or manipulative decision-making, aiming to stop failures before they reach production systems.
Prepared by Jonathan Pierce and reviewed by editorial team.
A medida que la IA se vuelve más frecuente en las empresas, los agentes de IA descontrolados representan una amenaza real. Pueden hacer un uso indebido de los datos, agotar recursos o tomar decisiones perjudiciales. La plataforma de Tenet Security podría ayudar a prevenir estos problemas, salvaguardando la infraestructura digital en la que confía a diario.
Tenet Security está dando un paso al frente para abordar el creciente riesgo de IA deshonesta en entornos empresariales. Su enfoque podría cambiar las reglas del juego en la seguridad de la IA, salvando potencialmente a las empresas de percances costosos y perjudiciales. Vale la pena reenviarlo si conoce a alguien en la industria tecnológica.
No especificado en el origen.
No especificado en el origen.
No left-leaning sources found for this story.
No right-leaning sources found for this story.
Comments