Guardrails

Guardrails refer to safety protocols embedded into AI systems to prevent vulnerabilities, ensure output integrity, and mitigate risks like prompt injection.