Taxonomy of Prompt Injection Attack Vectors
Researchers map 47 distinct prompt injection techniques across three dimensions of LLM attacks.
Priya Nandakumar
Senior Staff Writer
Priya Nandakumar
Priya spent eight years as a security architect at a mid-sized financial services firm before pivoting to technical journalism in 2017, where she has focused on identity infrastructure, access control standards, and the policy frameworks enterprises use to govern AI systems at scale.
16 stories
AI Agent Observability in Production Pipelines
Tracing every tool call and decision reveals why agents fail, not just that they do.
Priya Nandakumar
AI Acceptable Use Policies in Multi-Team Enterprises
Enterprises need layered AI policies, not one document, to prevent shadow AI breaches.
Priya Nandakumar
Building the Business Case for Centralized AI Governance
Ungoverned AI costs more over time, and centralized governance is how you prove it to the CFO.
Priya Nandakumar
AI Risk Management Frameworks for Enterprise Security Teams
Frameworks define AI risks but leave enforcement to the builder.
Priya Nandakumar
MCP Protocol Versioning and Backward Compatibility Management
Date-based versioning and runtime negotiation let modern and legacy MCP coexist—until they don't.
Priya Nandakumar
MCP Server vs AI Agent Responsibilities
Clearing up who's responsible when an AI agent goes rogue.
Priya Nandakumar
platforms for managing machine identity and access credentials in agentic AI workflows
Enterprises need purpose-built platforms to manage agent credentials safely.
Priya Nandakumar
approaches to zero-trust access enforcement for AI agents interacting with enterprise APIs
Enterprises need four governance layers, not one tool, to secure AI agents accessing APIs.
Priya Nandakumar
Governing MCP Tool Permissions at the Server Level
Enterprises must enforce tool permissions at the server level, not rely on the protocol.
Priya Nandakumar
Access Policy Enforcement for AI Tools Across Business Units
Enterprise AI policies fail because enforcement hasn't followed tools into CRMs, code, and APIs.
Priya Nandakumar
Non-Human Identity Management for AI Agents
Workload identity tokens and least-privilege scoping close the governance gap agents create.
Priya Nandakumar
Least Privilege Enforcement for Agentic AI Workflows
Static grants let agent access creep until a breach or audit finally catches it.
Priya Nandakumar
OAuth and SPIFFE for AI Agent Identity
SPIFFE proves what agents are; OAuth delegates what they can do.
Priya Nandakumar
IGA vs IAM for AI Agent Populations
AI agents demand continuous runtime governance where humans need quarterly reviews.
Priya Nandakumar
Enterprise AI Governance Framework Design and Ownership
Governance requires architecture, accountability, and enforcement operating together at every stage.
Priya Nandakumar