# AI Runtime Security Tools > Sixteen AI runtime security tools scored on what they stop before it completes, where they enforce, latency, deployment and proof. Read 16 Nov 2026. Every page has a markdown twin: add index.md to its address, or send Accept: text/markdown. Full text of every page: https://airuntimesecuritytools.com/llms-full.txt ## home - [AI Runtime Security Tools, Ranked by Action Stopped](https://airuntimesecuritytools.com/): Compare 16 AI runtime security tools by the harmful actions each stops before it completes. Pick an action, see the tools, read the scores. ## ranking - [AI Runtime Security Tools Ranking, 16 Nov 2026](https://airuntimesecuritytools.com/ranking/): Full ranking of 16 AI runtime security tools with totals out of 100, seven weighted criteria and the evidence read on 16 Nov 2026. ## method - [How We Score AI Runtime Security Tools](https://airuntimesecuritytools.com/method/): How AI runtime security tools are ranked: one research plan, 896 dated facts in 16 files with quotes checked, blind scoring, weights set first. ## vendors-index - [AI Runtime Security Vendors Compared](https://airuntimesecuritytools.com/vendors/): Profiles of 16 AI runtime security vendors: what each stops before it completes, where it enforces, how it deploys, with scores and sources. ## vendor - [Aim Security AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/aim-security/): What Aim Security stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Bay AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/bay/): What Bay stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Check Point AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/check-point/): What Check Point stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Cisco AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/cisco/): What Cisco stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Cloudflare AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/cloudflare/): What Cloudflare stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [CrowdStrike AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/crowdstrike/): What CrowdStrike stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [HiddenLayer AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/hiddenlayer/): What HiddenLayer stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Lasso Security AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/lasso-security/): What Lasso Security stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Microsoft AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/microsoft/): What Microsoft stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Noma Security AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/noma-security/): What Noma Security stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Operant AI AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/operant-ai/): What Operant AI stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Palo Alto Networks AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/palo-alto-networks/): What Palo Alto Networks stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Pillar Security AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/pillar-security/): What Pillar Security stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [SentinelOne (Prompt Security) AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/sentinelone/): What SentinelOne (Prompt Security) stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Straiker AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/straiker/): What Straiker stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. - [Zenity AI Runtime Security Profile](https://airuntimesecuritytools.com/vendors/zenity/): What Zenity stops at runtime before it completes, where it enforces, how it deploys and what is published, with scores and sources. ## bestfor-index - [Best AI Runtime Security Tool by Situation](https://airuntimesecuritytools.com/best-for/): Five situations re-weight the same scores: coding agents, customer LLM apps, SaaS agents, regulated response and fast rollout. ## bestfor - [Best AI Runtime Security Tool: Coding Agents](https://airuntimesecuritytools.com/best-for/coding-agents/): Which AI runtime security tool fits developers running coding agents: re-weighted scores for 16 Nov 2026 with the reason the weights change. - [Best AI Runtime Security Tool: Customer LLM App](https://airuntimesecuritytools.com/best-for/customer-llm-app/): Which AI runtime security tool fits a customer-facing LLM app: re-weighted scores for 16 Nov 2026 with the reason the weights change. - [Best AI Runtime Security Tool: SaaS Agents](https://airuntimesecuritytools.com/best-for/saas-agents/): Which AI runtime security tool fits agents inside SaaS workspaces: re-weighted scores for 16 Nov 2026 with the reason the weights change. - [Best AI Runtime Security Tool: Regulated Response](https://airuntimesecuritytools.com/best-for/regulated-response/): Which AI runtime security tool fits regulated response and audit: re-weighted scores for 16 Nov 2026 with the reason the weights change. - [Best AI Runtime Security Tool: Fast Rollout](https://airuntimesecuritytools.com/best-for/fast-rollout/): Which AI runtime security tool fits a fast rollout this quarter: re-weighted scores for 16 Nov 2026 with the reason the weights change. ## asset - [Runtime Action Type Matrix: What Each Tool Stops](https://airuntimesecuritytools.com/runtime-action-type-matrix/): Matrix of 16 AI runtime security tools by the action they stop before it completes: shell commands, files, pastes, installs, network calls and more. - [Enforcement Point Map: Where Each Tool Stops Harm](https://airuntimesecuritytools.com/enforcement-point-map/): Map of 16 AI runtime security tools by enforcement layer: model, application, agent and endpoint, with the evidence read on 16 Nov 2026. ## asset-method - [Runtime Action Type Matrix Methodology](https://airuntimesecuritytools.com/runtime-action-type-matrix/methodology/): How the Runtime Action Type Matrix was built: the 8 actions, the three levels, the evidence rules, the read date of 16 Nov 2026, and how to cite the data. - [Enforcement Point Map Methodology](https://airuntimesecuritytools.com/enforcement-point-map/methodology/): How the Enforcement Point Map was built: the four layers, the three levels, the evidence rules, the read date of 16 Nov 2026, and how to cite the data. ## asset-dim - [Prompt Injection Blocking: AI Runtime Security Tools](https://airuntimesecuritytools.com/runtime-action-type-matrix/prompt-injection/): Which AI runtime security tools block prompt injection, where the check sits in the stack, and what published research says about limits. - [Harmful Output Blocking: AI Runtime Security Tools](https://airuntimesecuritytools.com/runtime-action-type-matrix/harmful-output/): Which AI runtime security tools block harmful or data-leaking model output before it is shown or used, and where output controls sit in the stack. - [Model Layer Enforcement: AI Runtime Security Tools](https://airuntimesecuritytools.com/enforcement-point-map/model-layer/): What an AI runtime security control at the model layer sees and misses, how it is deployed (SDK, proxy, API), and which vendors document it. - [Application Layer Enforcement: AI Runtime Security Tools](https://airuntimesecuritytools.com/enforcement-point-map/application-layer/): What an AI runtime security control at the application layer sees and misses, how it is deployed (SDK, proxy, API), and which vendors document it. - [Agent Layer Enforcement: AI Runtime Security Tools](https://airuntimesecuritytools.com/enforcement-point-map/agent-layer/): What an AI runtime security control at the agent layer sees and misses, how it hooks tool calls and plans, and which vendors document it. - [Endpoint Layer Enforcement: AI Runtime Security Tools](https://airuntimesecuritytools.com/enforcement-point-map/endpoint-layer/): What an AI runtime security control at the endpoint layer sees and misses, how an endpoint agent or sensor is deployed, and which vendors document it. ## datasets-index - [AI Runtime Security Datasets](https://airuntimesecuritytools.com/datasets/): Download the Runtime Action Type Matrix and Enforcement Point Map as CSV and JSON, with methods, cite line and CC BY 4.0 license. ## dataset - [Runtime Action Type Matrix dataset, CSV and JSON](https://airuntimesecuritytools.com/datasets/runtime-action-type-matrix/): Download the Runtime Action Type Matrix from AI Runtime Security Tools as CSV or JSON under CC BY 4.0, with the cite line and the date read. - [Enforcement Point Map dataset, CSV and JSON](https://airuntimesecuritytools.com/datasets/enforcement-point-map/): Download the Enforcement Point Map from AI Runtime Security Tools as CSV or JSON under CC BY 4.0, with the cite line and the date read. ## guides-index - [AI Runtime Security Guides](https://airuntimesecuritytools.com/guides/): Twelve guides on AI runtime security: inline versus out-of-band, fail-open, latency, logging, RAG, metrics, cloud controls and open source. ## guide - [AI Runtime Security Explained](https://airuntimesecuritytools.com/guides/ai-runtime-security-explained/): What AI runtime security is, which actions it stops, where it enforces and how it differs from posture management, with sources read on 16 Nov 2026. - [AI Runtime Security vs Guardrails vs Red Teaming](https://airuntimesecuritytools.com/guides/ai-runtime-security-vs-guardrails-vs-red-teaming/): How AI runtime security, guardrails and red teaming differ: when each acts, what each produces and how they fit together, with sources. - [AIDR vs EDR vs XDR: What Each Watches](https://airuntimesecuritytools.com/guides/aidr-vs-edr-vs-xdr/): What AI detection and response (AIDR), EDR and XDR each watch, and how to run them together, with vendor documentation read on 16 Nov 2026. - [Cloud Provider Built-In Runtime Controls](https://airuntimesecuritytools.com/guides/cloud-provider-built-in-runtime-controls/): What Amazon Bedrock Guardrails, Azure AI Content Safety and Google Model Armor include, and when a separate AI runtime security tool is needed. - [Fail Open or Fail Closed: What Vendors Publish](https://airuntimesecuritytools.com/guides/fail-open-or-fail-closed-what-vendors-publish/): What happens when an AI runtime control fails, which vendors document a failure mode, and what to ask for, using documents read on 16 Nov 2026. - [Inline or Out-of-Band Runtime Enforcement](https://airuntimesecuritytools.com/guides/inline-or-out-of-band-runtime-enforcement/): Inline enforcement can stop an AI action before it completes; out-of-band monitoring sees it afterward. How to choose, with vendor and cloud sources. - [Open-Source AI Runtime Security Projects](https://airuntimesecuritytools.com/guides/open-source-ai-runtime-security-projects/): Open-source projects that enforce at runtime: NeMo Guardrails, Llama Guard, Prompt Guard, LLM Guard and Guardrails AI, compared with commercial tools. - [Runtime Protection for Customer-Facing LLM Apps](https://airuntimesecuritytools.com/guides/runtime-protection-for-customer-facing-llm-apps/): Attacks on a public LLM app and the controls that sit in the request path: input checks, output checks, data filters and logs, with sources. - [Runtime Security for RAG Pipelines](https://airuntimesecuritytools.com/guides/runtime-security-for-rag-pipelines/): What goes wrong in a retrieval-augmented generation pipeline at runtime and where to place controls: ingest, retrieval, prompt, output. - [Runtime Security Latency Budgets for LLM Apps](https://airuntimesecuritytools.com/guides/runtime-security-latency-budgets-for-llm-apps/): How much delay an AI runtime control may add to an LLM app, how vendors state it, and how to measure it yourself, with sources read on 16 Nov 2026. - [Runtime Security Metrics: False Positives and Blocks](https://airuntimesecuritytools.com/guides/runtime-security-metrics-false-positives-and-blocked-rate/): Which metrics show whether an AI runtime control works: blocked rate, false positives, misses and latency, and how to set a false positive budget. - [What to Log From Runtime Security Decisions](https://airuntimesecuritytools.com/guides/what-to-log-from-runtime-security-decisions/): Which AI runtime security decisions to record, what fields to keep, and how to send them to a SIEM, with vendor documentation read on 16 Nov 2026. ## glossary-index - [AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/): Twenty-eight plain definitions for AI runtime security: fail-open, inline enforcement, decision latency, output filtering, sidecars and more. ## term - [What Is Anomaly Detection? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/anomaly-detection/): Anomaly Detection defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Behavioral Baseline? Glossary](https://airuntimesecuritytools.com/glossary/behavioral-baseline/): Behavioral Baseline defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Block vs Detect? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/block-vs-detect/): Block vs Detect defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Canary Token? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/canary-token/): Canary Token defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Decision Latency? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/decision-latency/): Decision Latency defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is eBPF? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/ebpf/): eBPF defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Fail Closed? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/fail-closed/): Fail Closed defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Fail Open? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/fail-open/): Fail Open defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is False Positive Budget? Glossary](https://airuntimesecuritytools.com/glossary/false-positive-budget/): False Positive Budget defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Inline Enforcement? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/inline-enforcement/): Inline Enforcement defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Input Filtering? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/input-filtering/): Input Filtering defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Jailbreak? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/jailbreak/): Jailbreak defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Kernel-Level Monitoring? Glossary](https://airuntimesecuritytools.com/glossary/kernel-level-monitoring/): Kernel-Level Monitoring defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Model Denial of Service? Glossary](https://airuntimesecuritytools.com/glossary/model-denial-of-service/): Model Denial of Service defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Output Filtering? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/output-filtering/): Output Filtering defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is PII Redaction? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/pii-redaction/): PII Redaction defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Rate Limiting? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/rate-limiting/): Rate Limiting defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Reverse Proxy? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/reverse-proxy/): Reverse Proxy defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Runtime Protection? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/runtime-protection/): Runtime Protection defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is SDK Integration? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/sdk-integration/): SDK Integration defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Sensitive Information Leakage? Glossary](https://airuntimesecuritytools.com/glossary/sensitive-information-leakage/): Sensitive information leakage defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Sidecar Deployment? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/sidecar-deployment/): Sidecar Deployment defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is SIEM Export? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/siem-export/): SIEM Export defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is SOAR Playbook? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/soar-playbook/): SOAR Playbook defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is System Prompt Leakage? Glossary](https://airuntimesecuritytools.com/glossary/system-prompt-leakage/): System Prompt Leakage defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Telemetry? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/telemetry/): Telemetry defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Token Budget? AI Runtime Security Glossary](https://airuntimesecuritytools.com/glossary/token-budget/): Token Budget defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. - [What Is Tool Call Authorization? Glossary](https://airuntimesecuritytools.com/glossary/tool-call-authorization/): Tool Call Authorization defined for AI runtime security buyers, with one sourced example, read on 16 Nov 2026. ## news-index - [AI Runtime Security News](https://airuntimesecuritytools.com/news/): Dated items on AI runtime security releases, acquisitions, attack research against guardrails and cloud provider control changes. ## news - [Cisco open sources Project Quarry agent detections](https://airuntimesecuritytools.com/news/cisco-project-quarry-agent-detections/): Cisco open sourced Project Quarry on 28 Oct 2026, a detection language for multi-step agent attacks that maps findings to block, redact or log. - [Cisco DefenseClaw pairs with NVIDIA OpenShell sandbox](https://airuntimesecuritytools.com/news/cisco-defenseclaw-openshell-coding-agents/): Cisco's 21 Oct 2026 post describes DefenseClaw checking prompts, tool calls and results in an NVIDIA OpenShell sandbox for enterprise coding agents. - [Bedrock Guardrails grounding checks add Standard tier](https://airuntimesecuritytools.com/news/bedrock-guardrails-grounding-standard-tier/): Amazon Bedrock's document history records on 19 Oct 2026 that Guardrails contextual grounding checks now support the Standard safeguard tier. - [AgentCorruption: AWS makes IMDSv2 default in AgentCore](https://airuntimesecuritytools.com/news/zenity-agentcorruption-aws-agentcore-fix/): Zenity Labs disclosed a chain of AWS AgentCore flaws on 8 Oct 2026; AWS made IMDSv2 the default and reduced the default execution role's permissions. - [BRANCH paper bypasses multi-scanner AI guardrails](https://airuntimesecuritytools.com/news/branch-bypasses-multi-scanner-guardrails/): An arXiv paper reports a search method that beat six multi-scanner guardrails in tests and transferred to eight commercial ones. - [Noma extends agent runtime protection to the endpoint](https://airuntimesecuritytools.com/news/noma-endpoint-agent-security/): Noma announced on 28 Sep 2026 endpoint discovery, access control and AI-DR blocking for agents, MCP servers and skills on employee machines. - [EvasionBench: agents try to evade runtime monitors](https://airuntimesecuritytools.com/news/evasionbench-agents-evade-runtime-monitors/): An arXiv paper reports LLM agents circumventing runtime monitors under ordinary task pressure, with success up to 88% in best-of-3 tests. - [Google renames AI Protection findings tied to Model Armor](https://airuntimesecuritytools.com/news/google-ai-protection-model-armor-finding-names/): Google's Security Command Center release notes of 24 Sep 2026 rename AI Protection finding categories, including those for tuned models and Model Armor. - [Study flips Prompt Guard 2 verdicts with paraphrases](https://airuntimesecuritytools.com/news/prompt-guard-2-explainability-bypass-study/): An arXiv case study of Prompt Guard 2 found that synonym swaps and paraphrasing flip classifier verdicts, and that explanations lower the cost of bypass. - [HiddenLayer raises $100M Series B for agent runtime](https://airuntimesecuritytools.com/news/hiddenlayer-100m-series-b/): HiddenLayer announced a $100 million Series B on 2 Sep 2026 and will fund Agentic Runtime Security and a new Agent Harness Security product. - [Operant launches Semantic Firewall for agent actions](https://airuntimesecuritytools.com/news/operant-semantic-firewall-launch/): Operant AI shipped Semantic Firewall on 27 Aug 2026 with Token Meter, Browser AI Protection and wider Claude coverage for inline agent control. - [Prompt Security research on agent approval prompts](https://airuntimesecuritytools.com/news/prompt-security-agent-approval-prompts/): Prompt Security's research on instruction files, marketplace skills and approval prompts shows why prompts must display the real destination and payload. ## faq - [AI Runtime Security FAQ](https://airuntimesecuritytools.com/faq/): Answers on AI runtime security and AIDR: what they are, how they differ from guardrails and EDR, latency, failure modes and open source. ## about - [About AI Runtime Security Tools](https://airuntimesecuritytools.com/about/): What this site publishes: a ranked comparison of AI runtime security tools, two datasets, guides, a glossary and news, built from dated evidence. ## editorial-policy - [Editorial Policy](https://airuntimesecuritytools.com/editorial-policy/): How facts are gathered, how scores are set, how vendor claims are worded and how changes are logged on AI Runtime Security Tools. ## corrections - [Corrections and Tips](https://airuntimesecuritytools.com/corrections/): Report a factual error or send a tip about an AI runtime security tool. Corrections are logged on the method page with dates and sources. ## contact - [Contact AI Runtime Security Tools](https://airuntimesecuritytools.com/contact/): Contact the editor of AI Runtime Security Tools by email for corrections, tips, dataset questions and citation requests. ## privacy - [Privacy Policy](https://airuntimesecuritytools.com/privacy/): What data AI Runtime Security Tools collects: CloudFront access logs kept 90 days, no cookies, no analytics scripts and no accounts. ## terms - [Terms of Use, AI Runtime Security Tools](https://airuntimesecuritytools.com/terms/): Terms for using AI Runtime Security Tools: CC BY 4.0 datasets, attribution with the edition date, read dates and corrections. ## Data - [Runtime Action Type Matrix (CSV)](https://airuntimesecuritytools.com/datasets/runtime-action-type-matrix.csv) - [Runtime Action Type Matrix (JSON)](https://airuntimesecuritytools.com/datasets/runtime-action-type-matrix.json) - [Enforcement Point Map (CSV)](https://airuntimesecuritytools.com/datasets/enforcement-point-map.csv) - [Enforcement Point Map (JSON)](https://airuntimesecuritytools.com/datasets/enforcement-point-map.json) - [News feed](https://airuntimesecuritytools.com/feed.xml)