Skip to content

Add trust guardrails, AI disclaimers, and PII privacy notice #1011

Description

@jeromehardaway

Context

Neither the resume translator nor the career guides disclose that content is AI-generated. Cert coverage percentages are presented with progress bars that look authoritative. Salary figures are stale. Enrichment layer failures are invisible to users. Veterans are making career decisions based on this output without knowing its limitations.

Depends on: #1010 (LangGraph pipeline)

What

Add honest disclaimers, surface warnings, constrain quantification, and add a privacy notice.

Tasks

Constrain rank-scope quantification

In the LangGraph Generate Translation node prompt:

Change from telling the AI to pick numbers from ranges to:

Present scope as ranges (e.g., "4-10 team members", "$100K-$2M in assets").
Do NOT pick a single number from a range — keep the range to maintain accuracy.

Surface enrichment warnings to users

File: src/components/translator/TranslatorResults.tsx

  • Read _warnings from API response
  • Show yellow banner: "Some enrichment data was unavailable for this MOS code. Results may be less detailed."

Add AI disclosure to translator results

File: src/components/translator/TranslatorResults.tsx

  • After the green "AI-enhanced translation complete" banner, add:

    "This translation was generated by AI. Review all content carefully — verify metrics, certifications, and job titles against your actual experience before using on a resume."

Add cert disclaimer

File: src/components/translator/CertPathwaysCard.tsx

  • After "Based on your MOS training" text, add:

    "Certification readiness is an AI estimate. Check each certification body's official prerequisites before registering for exams."

Add career guide disclaimers

File: src/pages/career-guides/[mos].tsx

  • Add disclaimer banner after header section:

    "This career guide was generated using AI analysis of military job code data. Salary figures, certification coverage percentages, and career matches are estimates — verify with official sources before making career decisions."

  • Change "Ready to Certify" → "Potentially Ready to Certify"
  • Add "(estimated)" to ACE credits display
  • Add data freshness footer: "Career guide data last generated: [date]"

Add privacy notice to translator form

File: src/components/translator/TranslatorForm.tsx

  • Near submit button:

    "Your information is processed securely. Personal details (names, SSNs, contact info) are automatically redacted before being sent to our AI service."

Output validation flags in UI

File: src/components/translator/TranslatorResults.tsx

  • If _validationFlags exist on a bullet (from the Validate Output graph node), show a small warning icon next to that bullet with tooltip explaining the concern

Verification

  • Translate E-5 11B → verify bullets show ranges, not point estimates
  • Use MOS with sparse data → verify yellow warning banner
  • Check translator results page → AI disclaimer visible
  • Check cert card → cert disclaimer visible
  • Visit /career-guides/11b → disclaimer banner, "Potentially Ready to Certify", freshness date
  • Check translator form → privacy notice near submit button

Files

  • src/components/translator/TranslatorResults.tsx
  • src/components/translator/CertPathwaysCard.tsx
  • src/components/translator/TranslatorForm.tsx
  • src/pages/career-guides/[mos].tsx
  • src/lib/translation-graph.ts (prompt changes)

Metadata

Metadata

Labels

Type

No type

Fields

No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions