AI GatewayUpdated 2026-01-13
Policy-as-code for LLM behavior
Define LLM behavior with policy-as-code rules, allowlists, denylists, and reason codes in an AI gateway.
Policy-as-code makes LLM behavior explicit and testable, just like infrastructure-as-code.
Policy Gateway is an AI gateway that turns LLM guardrails into versioned JSON with rollouts and audit logs.
Definition
Policy-as-code for LLM behavior
Policy-as-code for LLM behavior is the practice of defining LLM rules in versioned, reviewable JSON so outcomes are predictable and auditable.
Why it matters
- Policies are versioned and can be reviewed like code.
- Reason codes make moderation outcomes deterministic.
- Rollouts and simulations reduce risk during policy changes.
How it works
- 01Define allowlist and denylist topics in policy JSON.
- 02Add flagged categories and response patterns for safe outputs.
- 03Apply the policy via the AI gateway endpoint with policy_id.
Runnable cURL snippet
curl https://api.abliteration.ai/policy/chat/completions \
-H "Authorization: Bearer $POLICY_KEY" \
-H "Content-Type: application/json" \
-H "X-Policy-User: user-4481" \
-H "X-Policy-Project: knowledge-bot" \
-d '{
"model": "abliterated-model",
"messages": [{"role":"user","content":"Share our internal incident response steps."}],
"policy_id": "policy-as-code"
}'Example policy JSON
{
"policy_id": "policy-as-code",
"name": "Policy as code",
"owner": "Security team",
"description": "Explicit allowlist and denylist behavior.",
"rules": {
"allowlist": ["public docs", "status page"],
"denylist": ["internal incident", "credential sharing", "security bypass"],
"flagged_categories": ["self-harm/intent", "violence/graphic", "sexual/minors"],
"response_pattern": "refuse",
"rewrite_instead_of_refuse": false,
"redact": true,
"reason_codes": ["ALLOW", "REFUSE", "REDACT"]
},
"org_controls": {
"project_keys": true,
"user_quotas": true,
"audit_logs": true,
"data_classification": "restricted",
"user_quota": { "requests": 100, "tokens": 10000, "window": "daily" },
"project_quota": { "requests": 5000, "tokens": 500000, "window": "monthly" }
},
"rollout": {
"shadow": { "enabled": true, "sample_percent": 25, "targets": ["knowledge-bot"] },
"canary": { "enabled": true, "sample_percent": 10, "targets": ["knowledge-bot"] },
"rollback_on_spike": true
},
"refusal_replacement": { "mode": "refuse", "escalation_path": "security@company.com" }
}Before and after
Before (no policy-as-code)
Assistant shares internal steps without a clear rule set.
After (policy-as-code)
decision: refuse reason_code: REFUSE denylist_hits: ["internal incident"]
Run the Policy Gateway simulator
Validate policy-as-code rules before rolling them into production.
Run a simulationFAQ
Frequently asked questions.
How do I version policies?
Export policy JSON from the console and store it in Git with your application code.
Can I canary policy changes?
Yes. Use rollout.canary to apply changes to a percentage of traffic first.