AI Agents Are Going Rogue: Here's How to Rein Them In
Human-centered identity frameworks are incorrectly being applied to AI agents, creating the potential for catastrophe at machine speed, Poghosyan argues.
Explore the latest frameworks in information security. Stay updated on guidelines to protect your digital assets and ensure data privacy.
Search across headline titles and summaries.
Background for this topic.
A security framework is an organized set of principles, practices, and controls for managing information and technology risk. Frameworks such as the NIST Cybersecurity Framework, ISO/IEC 27001, and COBIT help organizations structure activities including identifying assets and risks, protecting systems, detecting events, responding to incidents, and recovering operations. They are reference models rather than automatically effective security programs: an organization must select and implement measures appropriate to its systems, threats, and risk tolerance.
Practitioners use frameworks to assign responsibilities, prioritize vulnerability remediation, assess suppliers and cloud services, and document why particular controls are in place. They also provide a common vocabulary for audits, regulatory or contractual evidence, and measuring improvement over time. News under this tag may concern revisions to framework requirements, mappings between frameworks, assessment findings, or failures caused by treating a framework checklist as proof that controls work. A framework can guide governance and security operations, but it does not replace technical testing, continuous monitoring, or judgment about specific attack surfaces.
Human-centered identity frameworks are incorrectly being applied to AI agents, creating the potential for catastrophe at machine speed, Poghosyan argues.
CISOs must navigate five critical dimensions of AI in cybersecurity: augmenting security with AI, automating security with AI, protecting AI systems, defending against AI-powered threats, and aligning AI strategies with business goals. Neglecting any of these areas is a recipe for disaster.
AI's Impact on Productivity and Employment Demands Proactive Policy ActionThe future of work is no longer speculative, it is already being coded. From automated writing assistants to robotic warehouse employees, artificial intelligence is entering every part of the modern workforce faster than regulations or social frameworks can adapt.
Reconnaissance and BEC are among the malicious activities attackers commit after compromising cloud accounts, using a framework based on the TruffleHog tool.