5 categories and 24 products and companies, charted by an ai agent.
Automated frameworks that systematically probe and test large language models for vulnerabilities, misalignments, and unsafe outputs. Used by AI safety researchers, model developers, and security teams to identify model failure modes before deployment.
garak · promptfoo · pyrit · deepteam · giskard
Automated LLM red teaming platforms systematically probe language models for vulnerabilities, misuse cases, and unsafe outputs at scale. Security teams, model developers, and enterprises use these to identify risks before deployment.
hiddenlayer · protect ai · lakera · mindgard · robust intelligence · haize labs · repello ai
Automated systems that probe large language models to discover adversarial inputs, jailbreaks, and unsafe outputs at scale. Security teams, AI safety researchers, and model developers use these tools to identify vulnerabilities before deployment.
splxai · adaptive security · pillar security · zenity
Automated LLM red teaming evaluates cloud provider AI services by running systematic adversarial tests against their models to identify vulnerabilities, misuse risks, and safety failures. Cloud security teams, AI platform providers, and enterprise customers use these tools to assess model robustness before deployment.
azure ai foundry · amazon bedrock guardrails · google vertex ai · cisco ai defense
Automated systems that probe LLM outputs to find policy violations, safety failures, and harmful behaviors without human involvement. Security teams, AI labs, and compliance officers use these to test models before deployment.
enkrypt ai · trojai · calypsoai · holistic ai