AI Agent Security
Governing Salesforce Agentforce Deployments
Agentforce agents run as a dedicated Salesforce agent user, not the invoking human, so what that user's permission sets grant is the first thing to check.
AI Agent Security
AI Agent Security
Identity, authorization, guardrails, and trust for autonomous AI agents.
Start here
AI Agent Security: The Definitive Guide
AI agent security and agentic AI security explained: identity, authorization, guardrails, trust scoring, A2A, audit logging and incident response.
Read the guide →
AI Agent Security
Governing Amazon Bedrock Agents and Action Groups
How Bedrock Agents and action groups grant agents tool access, and why the Lambda functions behind them are an ungoverned surface most inventories miss.
AI Agent Security
AI Agent Scope Creep: Detection and Containment
Scope creep is an agent exceeding its intended boundary with no injected content and no over-broad grant. How to detect it and contain it before it happens.
AI Agent Security
Aikido Rebuilt the Gym Booking Hack: It Worked 9 of 10 Runs
Aikido Security reconstructed the Australian gym-booking hack. Claude Opus 4.6 on OpenClaw exploited the same flaw in 9 of 10 runs, 5 of them unprompted.
AI Agent Security
UK AISI Agent Incident Report: 19 Unsanctioned Actions
The UK AI Security Institute's own agents took 19 unsanctioned actions during cyber evaluations. What the report actually found, and what it argues for.
AI Agent Security
How a GitHub Issue Could Hijack a Google ADK Agent
Pillar Security found a GitHub issue could impersonate Google's own bot to trigger a privileged Google ADK workflow. Google deleted the affected workflows.
AI Agent Security
Hermes Agent's Threat Model: CVEs, Memory, and Skill Risk
Hermes Agent crossed 214,000 GitHub stars while carrying two disclosed CVEs and an unbounded memory attack surface. What a security review needs to know.
AI Agent Security
Hermes Agent vs. OpenClaw: The Governance Differences
Hermes Agent and OpenClaw scope identity, tool permissions, and memory differently. A governance-first comparison, not another feature table.
AI Agent Security
How to Run an AI Agent Security Incident Tabletop Exercise
A guide to designing and facilitating a tabletop exercise for an AI agent security incident: roles, injects, a 90-minute run sheet, and the debrief.
AI Agent Security
LangChain and LangGraph CVEs: What Deployments Need to Know
Three CVEs disclosed in March 2026 exposed files, secrets, and checkpoint data across LangChain and LangGraph. What they are and how to verify you're patched.
AI Agent Security
NemoClaw DNS Rebinding: One Webpage Could Rewrite the Model
NVIDIA NemoClaw shipped Ollama unauthenticated on 0.0.0.0:11434 — a DNS-rebinding attack from any webpage could rewrite a local agent's model permanently.
AI Agent Security
OpenAI Agents SDK Guardrails: What They Cover, and the Gaps
OpenAI Agents SDK guardrails validate input and output — but only for specific agents in a handoff chain. What the docs disclose, and what to add.
AI Agent Security
Prompt Injection vs. Jailbreaking: The Practical Difference
Prompt injection hits the application layer; jailbreaking hits the model's safety alignment. OWASP groups both under LLM01 — here's the practical difference.
AI Agent Security
Governing Windsurf Cascade: Agent Access to Your Repo
Windsurf's Cascade agent edits across files and runs terminal commands from inside the IDE. Here is what to scope before letting it loose on a real repository.
AI Agent Security
Cloud-Executed Coding Agents: Governing Replit Agent
Replit Agent builds, runs, and deploys apps in a hosted cloud workspace with its own stored secrets. Here is the blast radius and where control has to sit.
AI Agent Security
Governing GitHub Copilot Agents: Repository-Level Risk
Copilot's coding agent works issues autonomously and opens pull requests from a cloud sandbox. Here is the repository-level risk and how to scope it.
AI Agent Security
Governing Gemini CLI: Controls for Terminal-Resident Agents
Gemini CLI runs in the terminal with local file and shell access. Here is the enterprise control set for a coding agent that lives outside the IDE.
AI Agent Security
Governing Autonomous SWE Agents Like Devin
Autonomous SWE agents like Devin work tickets end to end in their own sandbox. Here is what changes when a coding agent needs no human per step.
AI Agent Security
Governing Cursor in an Enterprise Codebase
Cursor's agent mode edits files and runs terminal commands with full developer privileges. Here is the blast radius and the controls that actually contain it.
AI Agent Security
Governing Claude Code in Enterprise Environments
Claude Code runs as a terminal agent with file, shell, and MCP access on the developer's machine. Here is how to scope and audit it at enterprise scale.
AI Agent Security
Governing Amazon Q Developer Agents
Amazon Q Developer's agentic features touch your IDE, CLI, and AWS account. Here is the enterprise control set for a coding agent wired into cloud identity.
AI Agent Security
Code Review Controls for AI-Generated Pull Requests
AI-generated pull requests arrive with plausible summaries and passing checks, which erodes review attention. Here is how to keep review rigorous.
AI Agent Security
Preventing Secrets Leakage Through AI Coding Agents
Coding agents read config files, run shell commands, and log tool arguments verbatim — three distinct paths for a secret to leak. Here is how to close each one.
AI Agent Security
Protect AI Alternatives: Evaluating AI Security Platforms
What Protect AI's AI/ML security posture category covers, where it stops, and how to evaluate alternatives for securing agents in production.
AI Agent Security
Agentic Commerce Fraud: Stolen Cards Through AI Agents
AI shopping agents create a new carding target: not the card, but the delegated payment authority. What HUMAN Security has documented, and what to instrument.
AI Agent Security
Agentic Browsers vs the Same-Origin Policy: The Atlas Update
UW researchers showed agentic browsers can undermine the same-origin policy. What OpenAI shipped for Atlas, and what it means for deployment decisions.
AI Agent Security
When the Attacker Was the Agent: OpenAI and Hugging Face
An OpenAI agent chained a weak API key into cluster admin on Hugging Face infrastructure in under 13 hours, with no human attacker involved.
AI Agent Security
MITRE ATLAS vs OWASP Agentic AI Top 10: Different Jobs
MITRE ATLAS and the OWASP Agentic AI Top 10 answer different questions. Here is which framework to reach for, and when a team needs both.
AI Agent Security
The Mexico Government AI Breach: A Defender's Postmortem
A single operator used Claude Code and GPT-4.1 to breach nine Mexican government agencies. What the incident shows defenders about agent-assisted offense.
AI Agent Security
ServiceNow Now Assist Agent Discovery Incident
How second-order prompt injection abused ServiceNow Now Assist's agent-to-agent discovery feature, and what it shows about multi-agent privilege design.
AI Agent Security
The Lethal Trifecta: A Framework for Agent Data Exfiltration
The lethal trifecta names the three properties that make prompt injection dangerous: private data, untrusted content, and external communication.
AI Agent Security
ForcedLeak and PipeLeak: Agentforce Breaches Explained
Two 2025-2026 prompt-injection disclosures in Salesforce Agentforce, ForcedLeak and PipeLeak, show how form data becomes a data-exfiltration path.
AI Agent Security
Agent Sandbox Escape: Containing Code Execution
A code interpreter runs attacker-influenceable code in your infrastructure. The agent sandbox escape paths that matter and how to build one that fails safe.
AI Agent Security
Insecure Output Handling: Agent Output as Injection Vector
Model output is untrusted input to whatever consumes it. How agent output causes XSS, SQL injection, command execution, and log poisoning, and how to bound it.
AI Agent Security
Are Agent Skills Safe? A SKILL.md Security Checklist
Review Agent Skills before installation: inspect scripts, pin versions, test runtime permissions, and record evidence with an editable review worksheet.
AI Agent Security
Securing AI Agents in CI/CD Pipelines
An agent in your CI/CD pipeline holds deploy credentials and reads attacker-submittable content. Why pull-request-triggered agents are the sharpest case.
AI Agent Security
WitnessAI Alternatives for Runtime AI Access Control
Evaluating alternatives to WitnessAI for runtime AI access control and compliance — what the category covers and how peers compare.
AI Agent Security
Voice Agent Security Risks
Voice agents add a biometric channel that can be synthesized, an audio channel that carries injection, and a real-time path with no room for review.
AI Agent Security
Noma Security Alternatives for Agentic AI Security
Evaluating alternatives to Noma Security for agentic AI security — red-teaming, agent containment, and where platform-bundled options fit.
AI Agent Security
Securing AI Coding Agents
AI coding agents read files, run tools, and push code autonomously. Learn the real risks — prompt injection, secret leakage — and how to contain them.
AI Agent Security
Securing Browser-Use Agents
A browser-use agent reads the open web and holds your session cookies: the highest-risk agent deployment shape in common use. Here is how to constrain it.
AI Agent Security
AI Agent Security for Startups
A pragmatic guide to AI agent security for startups: the controls that matter most when you are moving fast and have limited security resources.
AI Agent Security
The Confused Deputy Problem in AI Agents
AI agents are deputies acting for principals. Confuse which principal an agent serves and its permissions become the attacker's. How to structure authority.
AI Agent Security
Multi-Turn Jailbreaks: Why Per-Message Filtering Fails
Crescendo and other multi-turn jailbreaks spread the payload so no single message looks malicious. Why stateless filtering misses them and what catches them.
AI Agent Security
Excessive Agency: Controls for OWASP LLM06
Excessive agency turns a model mistake into an incident. Its three components (excessive functionality, permissions, and autonomy) and the controls for each.
AI Agent Security
Red Teaming AI Agents: A Practical Playbook
Red teaming AI agents is not prompt-hacking a chatbot: the objective is unauthorized action. A scope, a technique set, and a scoring model for useful findings.
AI Agent Security
RAG Pipeline Security: Vector Store Poisoning
A RAG pipeline pulls attacker-influenced content into a model's context. How vector stores get poisoned, how retrieval leaks across tenants, and what holds.
AI Agent Security
Runtime vs Design-Time Security for AI Agents
Why design-time controls are not enough for AI agents: the five runtime control classes that must sit outside the agent, and how they map to OWASP.
AI Agent Security
Continuous Adversarial Testing for AI Agents
How Praesidia continuously red-teams its own registered agents and MCP servers: opt-in, scope-guarded, synthetic payloads, no external egress.
AI Agent Security
Goal Hijacking in AI Agents (OWASP ASI01)
Goal hijacking redirects an AI agent's objective while it runs normally. How ASI01 attacks work and the controls that contain a hijacked goal.
AI Agent Security
AI Agent Memory Poisoning: Attacks and Defenses (ASI06)
Memory poisoning implants malicious content in an AI agent's persistent state. How ASI06 attacks persist across sessions and the defenses that work.
AI Agent Security
AI Agent Supply Chain Security (OWASP ASI04)
The AI agent supply chain — models, MCP servers, tools, prompts, frameworks — and the vetting, pinning, and attestation controls against ASI04.
AI Agent Security
Securing Inter-Agent Communication and A2A (ASI07)
Agent-to-agent channels are an attack surface: spoofed peers, injected payloads, unbounded delegation. Securing A2A against OWASP ASI07.
AI Agent Security
Trust Scores vs Allow-Lists for Agent Authorization
Static allow-lists gate identity; dynamic trust scores gate scope. Learn how each works, where each falls short, and why mature programs combine both.
AI Agent Security
Cascading Failures in Multi-Agent Systems (ASI08)
How one agent's error propagates through orchestrated systems — amplification, feedback loops, resource storms — and the ASI08 containment controls.
AI Agent Security
Rogue AI Agents and Kill Switches (OWASP ASI10)
Rogue agents run with valid credentials outside governance. The kill-switch architecture — containment tiers, revocation, drills — for OWASP ASI10.
AI Agent Security
A2A vs MCP: What's the Difference?
MCP connects agents to tools; A2A connects agents to each other. Understand how both protocols divide labor and what security controls each one demands.
AI Agent Security
How to Detect Prompt Injection: Signals That Work
A detection-focused how-to for prompt injection: pattern rules, ML classifiers, LLM judges, and behavioral signals — and where to place each one.
AI Agent Security
OWASP LLM Top 10 vs Agentic Top 10 for AI Agents
The OWASP LLM Top 10 read through an agent lens, how it differs from the Top 10 for Agentic Applications, and the controls that address each risk.
AI Agent Security
SSRF Risks in Agent Tool Calls
Any agent with a fetch tool is an SSRF primitive with a natural-language interface. The attack paths, why allow-lists beat denylists, and how to bound egress.
AI Agent Security
An AI Agent Incident Readiness Checklist
A practical checklist for AI agent incident readiness: inventory, instant revocation, tamper-evident audit trails, runbooks, and communication templates.
AI Agent Security
Securing the AI Agent Supply Chain
Provenance, attestation, and runtime verification protect AI deployments from compromised third-party agents and tools — and how supply chain security works.
AI Agent Security
Threat Model: Indirect Prompt Injection
Indirect prompt injection hijacks tool-using AI agents through poisoned external content. Learn the attack vectors and layered controls that contain them.
AI Agent Security
Insider Threat in the Age of AI Agents
Agents give insiders bulk access at machine speed and attribution that stops at a service identity. What changes for insider threat, and the controls that help.
AI Agent Security
System Prompt Extraction: What Leaks and Why It Matters
System prompts leak. Treat them as public and the damage is bounded; treat them as secret and extraction becomes a breach. What attackers get, and how to plan.
AI Agent Security
Securing LLM Provider API Keys
A leaked LLM provider key can be billed against and can expose prompt traffic. How these keys leak, why per-agent keys are the wrong fix, and what a proxy buys.
AI Agent Security
Model Extraction and Inference Attacks Against Agents
Query access alone can extract model behavior, infer training data, and reconstruct context. Which of these threats are real for agents and what mitigates them.
AI Agent Security
Threat Model: Agent-to-Agent Delegation Abuse
When AI agents delegate tasks to each other, the delegation chain becomes an attack surface. How to threat-model and contain A2A delegation abuse.
AI Agent Security
Incident Response for AI Agent Breaches
A practical incident response runbook for AI agent breaches: contain damage, revoke credentials, investigate with tamper-evident audit trails, and recover.
AI Agent Security
Federated AI: Sharing Agents Without Sharing Data
How signed trust manifests and scoped admission controls let organizations share AI agents across boundaries without exposing data or credentials.
AI Agent Security
Trust Scoring Models for Autonomous Agents
How agent trust scoring models aggregate identity, behavior, and attestation signals into a runtime gate that controls what autonomous agents can do.
AI Agent Security
HiddenLayer Alternatives: Model Security vs Agent Runtime
HiddenLayer's model security and AI detection category explained, how it differs from agent runtime authorization, and how to evaluate alternatives.
AI Agent Security
Tool-Use Safety: Sandboxing Agent Actions
Contain what AI agent tools can do: tool-level scoping, allow-lists, dry-runs, and human approval gates for high-consequence irreversible actions.
AI Agent Security
CalypsoAI Alternatives for AI Security Teams
CalypsoAI's GenAI validation and inference-security category explained, the evaluation criteria that matter, and when to look at a different layer.
AI Agent Security
Secrets Management for AI Agents
Keep API keys and credentials out of agent prompts and source code. The four pillars of secrets management: storage, delivery, access control, and rotation.
AI Agent Security
Governing Agents Built on the Vercel AI SDK
The Vercel AI SDK standardizes model calls and tool invocation for web apps. Identity, authorization and spend controls are the layer around it.
AI Agent Security
Code-Writing Agents: The Risk Profile of smolagents Design
Code agents write and execute Python to act, instead of calling predefined tools. That design choice trades a smaller interface for a larger one.
AI Agent Security
Governing Pydantic AI Agents: Type Safety Is Not a Guardrail
Pydantic AI validates a model's output shape; authorizing the action it triggers is a separate layer. Where type safety ends and governance begins.
AI Agent Security
Governing Mastra Agents and TypeScript Agent Workflows
Mastra gives TypeScript teams agents, durable workflows and memory. Identity, authorization, and spend controls sit in the layer around it.
AI Agent Security
Governing Flowise and Langflow Visual Agent Builders
Flowise and Langflow turn LangChain-style chains into drag-and-drop flows. The governance gap is the same as code-first LangChain, plus a review gap.
AI Agent Security
How to Govern CrewAI Agents: Controls for Crew-Based Systems
CrewAI orchestrates role-based agent crews in-process. Here is what it controls, where the governance gap sits, and what to add externally.
AI Agent Security
Agent-to-Agent Protocols and Interoperability
How to design safe A2A interoperability: agent cards, secure discovery, scoped credentials, and cross-org trust — in under 8 minutes.
AI Agent Security
Data Exfiltration Risks in Agentic AI
Agentic AI creates novel data exfiltration paths via over-broad tool access, chatty outputs, and prompt injection. Learn how to contain each risk layer.
AI Agent Security
Prompt Injection: Threats and Defenses
Prompt injection hides malicious instructions in content AI agents process. How direct and indirect variants work, and what defenses reduce the risk.
AI Agent Security
Zero Trust for AI Agents
Zero trust for AI agents means verifying every identity, enforcing least-privilege policy at every hop, and using behavioral trust scores at runtime.
AI Agent Security
Why AI Infrastructure Needs a New Security Model
Traditional IAM secures human users, not AI agents making thousands of calls per minute. Why a connection-centric model is the right foundation.
AI Agent Security
What Is Context Window Poisoning?
Context window poisoning corrupts an agent's active reasoning session with false content that later steps treat as established fact.
AI Agent Security
What Is Agentic AI Security Posture Management?
Agentic AI security posture management continuously assesses agent identities, permissions, and guardrail coverage to find drift before attackers do.
AI Agent Security
What Is Agent Drift, and How Do You Detect It?
Agent drift is a gradual change in an AI agent's behavior over time. Learn the detection signals and controls that catch it before it becomes an incident.
AI Agent Security
Inside the Agent Management Console
Register, configure, version, and debug every AI agent in your fleet from a single governed control surface with full audit trails and per-agent access control.
AI Agent Security
Agent-to-Agent (A2A) Communication, Governed
Governed A2A communication makes every inter-agent call authenticated, scoped, and audited — with agent cards, least-privilege identity, and guardrails.
AI Agent Security
Governed Connections Between Agents and Resources
Turn implicit agent-to-resource links into policy-bound connections with rate limits, spend caps, trust gates, and guardrails enforced at dispatch.
AI Agent Security
Securely Sharing AI Agents Across Organizations
Share AI agents with partner organizations under explicit policies — request caps, expiry, instant revocation — without handing over credentials.
AI Agent Security
Cross-Org Agent Federation with Trust Manifests
How signed trust manifests let organizations share AI agents across boundaries without shared secrets—each delegation explicit, verifiable, and revocable.
AI Agent Security
Trust Scores and Attestations: Which Agents to Trust
Learn how agent trust scores combine behavioral signals, compliance state, and cryptographic attestations into an auditable dispatch gate for AI agents.