Compliance and Permitted Use

Narrativ is built to produce high-signal, permissioned outputs that enterprises can use with confidence. This page defines:

what you're allowed to do with Narrativ data products

what you're not allowed to do

how permissions, provenance, and auditability are enforced across delivery

INFO

This is a high-level summary of permitted use and compliance posture. Your contract (MSA/DPA + Order Form) is the source of truth.

What makes Narrativ "enterprise-safe"

Most data products in the market rely on inference, opaque sourcing, or unclear rights. Narrativ is designed to be different:

Permissioned where user data is involved: use is tied to explicit user authorization and purpose

Traceable provenance: inputs and processing are attributable (source class + timestamps + permission state)

Purpose limitation: products are delivered for defined use cases, not "anything you want"

Versioned methodologies: index definitions change with clarity, not silently

Auditable controls: permission changes, eligibility states, and delivery events are logged

TIP

For how permissions work technically, see Consent and Permissions.

Data classes and what you're receiving

Narrativ products may include outputs derived from:

Managed sources

Public/system-level data collected via external APIs and web scrapers.

User-permissioned data

Data derived from connected profiles and approved data classes, under explicit user consent.

In most cases, B2B customers receive signals and indices (aggregated and normalized outputs), not raw account-level exports.

Permitted use (default)

Unless your Order Form states otherwise, Narrativ data products are typically permitted for:

Internal analytics and monitoring - dashboards, trend monitoring, cohort analysis, benchmarking

Research and strategy - market intelligence, narrative tracking, competitive attention

Model development (where explicitly licensed) - training, fine-tuning, evaluation, and QA using approved datasets

Automation and agent workflows - real-time index feeds and alerts consumed by internal systems and agents

WARNING

If you need a use case outside this scope, see Requesting expanded use below.

Permitted use matrix (quick reference)

Product	Internal	Research	AI/ML	Agentic	Resell	Re-ID
Indices (time series)	✅	✅	🔐	✅	❌	❌
Signals (aggregated)	✅	✅	🔐	✅	❌	❌
Alerts / webhooks	✅	✅	⚠️	✅	❌	❌
Scheduled exports	✅	✅	🔐	⚠️	❌	❌
AI-ready datasets	✅	✅	✅	⚠️	❌	❌
Custom enterprise feeds	✅	✅	✅	✅	📝	❌

✅ Allowed · 🔐 Requires AI/ML license · 📝 Requires written agreement · ⚠️ Limited / conditional · ❌ Prohibited

Prohibited use (always)

To protect user rights, maintain compliance, and preserve dataset integrity, the following are always prohibited:

Re-identification and individual targeting

High-risk / regulated decisioning

Redistribution and resale

Abuse and surveillance

AI / ML permitted use (practical rules)

If your license includes AI/ML use, the intent is to enable modern workflows without turning Narrativ into an uncontrolled training firehose.

Allowed (when licensed)

training/fine-tuning/evaluating models using approved datasets and approved purposes

using live signal feeds and alerts as context for internal agents and automation

Not allowed

training models to identify individuals, infer private identity, or reconstruct user-level data

using Narrativ data to build datasets whose primary value is identity-level targeting

combining Narrativ data with external data in a way that attempts to bypass consent constraints

Customer obligations (what we expect from buyers)

To maintain a compliant and stable ecosystem, customers agree to:

Purpose limitation: use the product only for the licensed purposes

Access controls: enforce least privilege (only authorized users/systems)

Security controls: store and transmit data securely (industry-standard safeguards)

No onward sharing: do not share with third parties without explicit written authorization

No re-identification: do not attempt to identify users or reconstruct account-level data

How Narrativ enforces permitted use

Narrativ enforces permitted use through both contractual and technical controls:

Permission-aware processing: outputs are generated under defined consent states

Provenance + audit logs: delivery events, permission states, and methodology versions are traceable

Product scoping: datasets/feeds are provisioned to specific customers and use cases

Abuse monitoring: anomalous usage patterns can trigger review, throttling, or suspension

Revocation-forward: changes to permissions apply going forward (future processing/delivery)

Retention and deletion

Retention is governed by your Order Form / DPA and product type. In general:

customers should retain data only as long as needed for the licensed use case

upon termination, continued use is not permitted unless explicitly agreed

deletion requirements may apply depending on jurisdiction, product class, and contract terms

TIP

For user rights framing, see Legal & Regulatory Framework → Data Rights.

Requesting expanded use

If you want to use Narrativ outputs beyond the default scope (e.g., broader redistribution, new AI use cases, regulated decisioning), we can evaluate it.

Quick checklist (buyers)

Before production use, confirm:

Our Order Form explicitly covers your intended use case(s)

You understand which products are derived from user-permissioned data vs managed sources

You have access controls and secure storage in place

You will not attempt re-identification or redistribution

You have a plan for retention and deletion aligned with the contract

What makes Narrativ "enterprise-safe"#

Data classes and what you're receiving#

Permitted use (default)#

Permitted use matrix (quick reference)#

Prohibited use (always)#

AI / ML permitted use (practical rules)#

Customer obligations (what we expect from buyers)#

How Narrativ enforces permitted use#

Retention and deletion#

Requesting expanded use#

Quick checklist (buyers)#

What makes Narrativ "enterprise-safe"

Data classes and what you're receiving

Permitted use (default)

Permitted use matrix (quick reference)

Prohibited use (always)

AI / ML permitted use (practical rules)

Customer obligations (what we expect from buyers)

How Narrativ enforces permitted use

Retention and deletion

Requesting expanded use

Quick checklist (buyers)