v1.1.0 · Effective July 8, 2026

AI & Data Use Disclosure

Exactly what the AI does, what it does not do, where its data comes from, and where it can be wrong.

Cited, scored, and reviewable - by design.

Decision-Grade Analyses, Moat intelligence rows, and regulatory answers carry citations to their federal source records. Confidence is scored, not asserted. Nothing here predicts an award.

What the AI does

  • Reads publicly available federal procurement data from SAM.gov, USAspending.gov, FPDS.
  • Reads the Customer Content you submit to your own workspace.
  • Produces analyses, drafts, summaries, and recommendations. Decision-Grade Analyses, Moat intelligence rows, and regulatory answers carry citations to their federal source records; elsewhere, an answer that has no citable source says so rather than implying one.
  • Scores its own confidence so you can tell a strong inference from a weak one.

What the AI does not do

  • It does not predict, guarantee, or improve the probability of any contract award.
  • It does not provide legal, procurement-law, financial, accounting, tax, or investment advice.
  • It does not make decisions. Its output is a draft for a qualified human to review.
  • It is not trained on your data, and it never sees another customer's data.
  • It has no access to non-public government information, and it confers no advantage obtained from any.

Known limitations

  • Language models can generate plausible statements that are wrong. This is inherent to the technology, not a defect we have eliminated.
  • Public procurement data lags reality. Award records, incumbents, and expiration dates may be stale or incomplete at the source.
  • Confidence scores express the system's own assessment of evidentiary support. They are not probabilities of winning anything.
  • Where the system cannot find supporting evidence, it is designed to say so rather than to guess. Report any instance where it does not.

Human review is required

DSS Nexus produces decision support, not decisions. Outputs are drafts and analyses that a qualified human must review before any external use. DSS Nexus does not guarantee any award, win rate, contract, ranking, revenue, or procurement outcome, and does not provide legal, financial, accounting, investment, tax, or procurement-law advice.

Before any Service output is submitted to a government customer, incorporated into a proposal, or relied on for a pricing, teaming, or bid/no-bid decision, a qualified person must review it against the primary source.

Your controls

These are buttons, not promises. Each control below is in Account & Settings → Data & Privacy.

  • Export your workspace data at any time from Account & Settings. We email you a secure, expiring download link.
  • Delete your workspace data from Account & Settings. Deletion is scheduled, confirmed in writing, and cancelable during the grace period; we complete it within 30 days.
  • We do not train models on Customer Content, so there is no training opt-out to configure.
  • Ask a question about any of this at info@dssadvisorygroup.com.
Related

Other policies

Version history

What changed, and when

v1.1.0 · August 5, 2026 - Scopes the citation claim to the surfaces where citation is enforced: Decision-Grade Analyses, Moat intelligence rows, and regulatory answers. Effective on production publication of this revision.

v1.0.0 · July 8, 2026 - First published version. Replaces the prior placeholder pages. Adds worldwide privacy rights, international transfer mechanisms, a data processing addendum, a published subprocessor list, and a commitment never to train models on Customer Content.

Contact

Questions about this policy?

DeSouza Strategic Systems LLC d/b/a DSS Advisory Group. Privacy: info@dssadvisorygroup.com. Legal: info@dssadvisorygroup.com. Security: info@dssadvisorygroup.com. Last updated July 8, 2026.

AI & Data Use Disclosure | DSS Nexus