Guardrails & Safety

Model monitoring and firewall (Arthur Shield) for enterprise AI, focused on risk, bias and policy enforcement.

Inference-perimeter security platform with scanners and red-team agents guarding enterprise model traffic.

AI model validation and runtime guardrails productized from Robust Intelligence inside Cisco's security stack.

Aporia's drift detection and AI guardrails folded into the Coralogix observability platform as its AI research arm.

Open-source LLM vulnerability scanner that runs pre-built probes for jailbreaks, leakage and injection.

AI red teaming and safety testing platform producing adversarial test suites for LLM applications.

Open-source output-validation framework where composable validators enforce schemas, policies and safety constraints on LLM I/O.

Automated red-teaming ('haizing') that stress-tests LLM systems to find jailbreaks and failure modes before deployment.

Low-latency API guarding against prompt injection, data leakage and toxic content, backed by the Gandalf attack dataset.

Open-source input/output scanner toolkit (35+ scanners) for PII, injection and toxicity checks on LLM traffic.

Automated AI red teaming platform testing LLMs, agents and multimodal models against MITRE ATLAS / OWASP-aligned attacks.

Programmable conversational guardrails toolkit using the Colang DSL, covering input, dialog, retrieval, execution and output rails.

AI security platform covering discovery, red teaming and runtime protection across the AI lifecycle.

Enterprise GenAI security platform monitoring employee and application LLM usage for injection, leakage and shadow AI.

Python Risk Identification Toolkit automating single- and multi-turn adversarial probing of GenAI systems.

AI red teaming platform whose ARTEMIS engine automates adversarial testing of LLM apps and agents.

Automated AI security testing and red teaming for AI assistants and agents from build to runtime.

AI-native security platform with red teaming and runtime guardrails for agentic applications.

Agent trust platform combining automated evaluation, red teaming and runtime defenses for AI agents.

Security and governance platform for enterprise AI agents and low-code copilots, including agent observability.