Data handling, plainly stated.

Visible only to org members. Bridger staff cannot read your cap statement without an explicit support ticket you initiate.

Same isolation. Multi-seat orgs share within the org boundary; no cross-org leakage.

Two-party DM threads. Counterparty sees the thread; nobody else does. Operator can audit message metadata (sender, recipient, timestamp) for abuse investigation but does not read message bodies absent legal process.

When the platform shows providers “X contractors are searching for Y capability,” that count is banded — providers see ranges (10-25, 25-50) not exact counts, and never account-attributable identities.

The CII (Country Instability Index) is computed from public sources and is the same number for every contractor. Not a privacy concern; included here for completeness.

Period. No data brokers. No advertising IDs. No commercial intelligence services.

No data partnerships, no “analytics products built on customer corpora,” no white-label intelligence resale.

The proposal-agent uses Anthropic Claude under zero-data-retention. Your prompts and completions are not retained by Anthropic for model improvement and never used to train any third-party model. See /legal/ai-use.

Without your in-writing authorization. Period.

Public — Brief content, public assessment summaries, CII scores. Indexed by search engines.

Internal — your cap statement, NAICS, capability narrative, target agencies. Visible to your org only.

Confidential — proposal drafts, message bodies, pipeline values, pricing assumptions. Visible to org members with confidential-tier access only.