> ## Documentation Index
> Fetch the complete documentation index at: https://docs.clausum.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Collaborative network — trust & enrollment
> Privacy guarantees and enrollment-first path for banks and PSPs joining the Clausum collaborative fraud network
Institutions often ask: *“Can we trust sharing signals with other banks?”* This page answers that for **legal, compliance, and fraud operations** — before any technical integration.
**Enrollment-first model:** you can **join and contribute** approved signal types **without** purchasing network consumption (**CLM-ADD-NET-C**). Consumption is an optional second phase under contract.
## Why enroll before consuming
| Without early contributors | With Clausum enrollment |
| - | - |
| Empty network — no sector value | Each institution adds value from day one |
| “Wait until others join” paralysis | **Contribute** is opt-in — no consume add-on required |
| Perceived PII exposure risk | **Granular publish policy** + peer anonymity |
| Months of delay | Scoped pilot: API + 3–5 legal-approved signal types |
**Commercial incentive:** eligible contributors may receive a **10 % discount** on **CLM-ADD-NET-C** (consumption), per contract.
***
## Clausum guarantees to your institution
| # | Guarantee | In practice |
| - | - | - |
| **G1** | **Peer anonymity** | Other participants never see your institution name — only alias `Participant INST-…`. Clausum retains attribution internally for governance. |
| **G2** | **Granular publish control** | You set **`contribute_list_types`** — e.g. only `national_id`, `bank_account`, `ip_range`. `null` = full catalog (28 types). |
| **G3** | **Granular consume control** (phase 2) | With **CLM-ADD-NET-C**, set **`consume_list_types`** and **`min_trust_score`** (0–100). |
| **G4** | **No internal note leakage** | `reason_detail` (case IDs, analyst notes) is **never** exposed cross-tenant. Consumers see **`reason_code`** only. |
| **G5** | **Sensitive data hashed** | National IDs, accounts, emails, phones, names, fingerprints — normalized/hashed for matching; safe labels in dashboard. |
| **G6** | **Operational revocation** | Revoke false positives in **Intelligence → My signals** — immediate for your org. |
| **G7** | **Trust governance** | Each contributor has a **trust score** (0–100) governed by Clausum staff. Consumers filter by threshold. |
| **G8** | **Assess fail-safe** | Network **augments** assess — local rules still apply if network is unavailable. |
Technical depth: [Intelligence network integration guide](/guides/clausum-intelligence-network) · 28 signal types catalog (engineering: `NETWORK_SIGNAL_CATALOG.md` in repo).
***
## What you control vs. what Clausum manages
| Setting | Owner | Where |
| - | - | - |
| Enable **contribution** | **Institution** | Dashboard → Intelligence → Settings |
| **Publish** signal types | **Institution** | `contribute_list_types` |
| **Consume** signal types (phase 2) | **Institution** | `consume_list_types` |
| Minimum contributor trust | **Institution** | `min_trust_score` |
| Enable **consumption** in assess | **Clausum** (contract **CLM-ADD-NET-C**) | Staff platform |
| Integration profile | **Clausum** | Based on `organization_type` |
| Contributor trust score | **Clausum** | Staff governance + audit |
***
## Recommended enrollment path (contribute first)
**CLM-MOD-INST** (Financial Hub) + Clausum onboarding — sandbox access.
Agree which signal types appear in **`contribute_list_types`** (start with 3–5 types).
Dashboard → **Contribute to network** · server key `clm_sk_*` or contributor key `clm_cnt_*`.
`POST /api/v1/network/intelligence` from sandbox — validate normalization and revocation runbook.
Contract **CLM-ADD-NET-C** · assess matches peer signals above your trust threshold.
***
## Signal catalog summary (28 types)
Legal can approve a subset for pilot. Groups:
| Group | Examples | Typical bank use |
| - | - | - |
| **Identity** | `national_id`, `customer_name`, `beneficiary_name` | Confirmed fraud on onboarding / transfers |
| **Contact** | `email`, `phone`, `email_domain` | Phishing, ATO |
| **Payment** | `bank_account`, `swift_bic`, `card_bin` | Mule accounts, SPEI, outbound fraud |
| **Device / network** | `ip_address`, `ip_range`, `device_fingerprint` | Card testing, botnets |
| **Location** | `billing_postal`, `address_hash` | Geo patterns |
| **Commerce** | `merchant_domain`, `submerchant_id`, `mcc` | Aggregator / PSP fraud |
***
## Mutual commitments (legal summary)
**Your institution commits to:**
* Publish only signals with agreed operational/legal basis (not experimental dumps).
* Stay within approved **`contribute_list_types`**.
* Maintain an internal **revocation** runbook for false positives.
* Use server-side keys only — never embed secrets in client apps.
* Name fraud and compliance contacts for the pilot.
**Clausum commits to:**
* Never disclose your institution name to other network participants.
* Never expose **`reason_detail`** cross-tenant.
* Apply catalog **normalization / hashing** rules.
* Enable immediate **revocation** of your signals.
* Operate contributor **trust score** and staff audit.
* Provide public docs, sandbox, and onboarding support (**CLM-SVC-ONB**).
Contractual details (DPA, retention, jurisdiction, SLA) are defined in the master agreement — this page describes **product capabilities**.
***
## Pre-signature checklist
* [ ] Approved signal types for **contribution** (explicit list)
* [ ] Internal legal basis for each category
* [ ] Named operational contact (revocation + escalation)
* [ ] Pilot on **sandbox.clausum.ai** before production
* [ ] (Phase 2) Decision on **CLM-ADD-NET-C** and initial `min_trust_score`
***
## Related
API, quick starts, troubleshooting
Financial Hub bundle
SPEI / disbursement gate
Permissions and key prefixes
Support: **[api@clausum.ai](mailto:api@clausum.ai)**
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.