Microsoft · Harm classifier · Last verified 2026-10-15

Azure AI Content Safety

Unverified

Harm classifiers with severity levels, Prompt Shields, and groundedness detection. Entra auth; Azure regions; API-versioned.

What it does

  • Harm classifiers with severity levels, Prompt Shields, and groundedness detection. Entra auth; Azure regions; API-versioned.

Pricing

  • Priced per text record (rates not re-checked). [NOT re-verified on 2026-10-15 — carried from prior research; check before relying on it]

Limits

  • Not recorded. [NOT re-verified on 2026-10-15 — carried from prior research; check before relying on it]

Status

  • Previously recorded as generally available. [NOT re-verified on 2026-10-15 — carried from prior research; check before relying on it]

Safety & fit notes

  • Good as a content gate; not a general decision model.
  • Same pattern applies: deterministic rules own hard constraints; the model is advisory; fail closed on errors.

Sources