GernalistYou ask. We investigate. The evidence is public.
Approved

Which AI note-taking and meeting-summary tools expose the biggest privacy tradeoffs?

Teams adopt meeting bots before reading retention and training policies.

0 votes0 evidence itemsAI productivity / privacy
Investigation brief

What this question needs before publication

Current state: this topic is approved. A public dossier only appears when the evidence is reviewable, claims are mapped to sources, and a human editor signs off on the final framing.

Trust showcase

Signals we will preserve

  • policy excerpts
  • last-reviewed timestamps
  • privacy tradeoff labels
What makes this publishable

Publishability checklist

1. Source floor

At least three independently reviewable sources before the question becomes a public claim set.

2. Claim mapping

Claims mapped to source IDs before publication, with caveats separated from supported findings.

3. Human review

Human editor review before a dossier goes public. AI output is never evidence.

How readers can help

Submit a source, correction, or firsthand context

Reader contributions stay in moderation until an editor can verify source quality, relationship disclosures, and whether the material belongs in the evidence ledger.

Review evidence intake
Share or save this question

Share context, not a verdict

Copy the page URL from your browser when you want to send the question to someone else. This page is the stable manual return path while evidence and status change.

When to revisit this question

What changes next

New source appears

Return when a new independently reviewable source appears and can change the source floor.

Status changes

Return when the status changes from proposed to researching, approved, or published.

Challenge gets stronger

Return when a claim can be challenged with better evidence, missing context, or a stronger counter-source.

What can change this status

Evidence changes before status changes

Independent sources

Independent sources can move a proposed question toward approved or researching when they make the question investigable.

Contradictory evidence

Contradictory evidence can keep a claim uncertain instead of publishable until editors separate supported findings from caveats.

Corrections and challenges

Corrections and challenges can reopen human review after publication when they change source confidence or claim wording.

Status changes still require human editorial review. This glossary does not move topics, publish dossiers, or run provider calls.

Moderated evidence submission

Add evidence for editor review

Submissions enter human moderation before they affect score, claims, or dossiers. This is a manual intake affordance only.

No AI extraction, search, crawling, or provider calls run from this formNo automatic source ledger changes
How this evidence intake area worksPublic submitter workflow
  1. 1. Prepare evidence quality with the checklist before typing into the form.
  2. 2. Check readiness and submit only after sign-in, profile sync, and topic state allow moderation intake.
  3. 3. Track owner-scoped status after submitted evidence appears in the signed-in status tracker.
  4. 4. Respond if editors ask for clarification from the signed-in clarification response surface.
  5. 5. Review seeded moderation examples to understand the manual queue shape.

No backend change, query change, mutation change, provider call, source-ledger mutation, automatic moderation, or publish action runs from this organization guide.

Evidence intake hierarchyScan order
Primary action: prepare and queue evidence

Start with the checklist, readiness guidance, and evidence fields.

Owner-only follow-up appears after sign-in

Status tracking and clarification responses stay scoped to the signed-in submitter.

Editorial examples stay last as read-only context

Seeded rows show the manual moderation shape after the active submitter workflow.

Visual hierarchy only; no backend, query, mutation, provider, source-ledger, automatic moderation, or publish behavior changes.

Orientation · Before you startCompact orientation · Start here before filling evidence fields

Public evidence intake guidance taxonomy and Public evidence intake guardrail summary stay together so submitters can scan phases and no-change boundaries before typing.

Phase labelsPresentation taxonomy

Prepare → Fields → Queue → Review → Follow-up. Use these labels to scan the existing guidance before queueing evidence.

Taxonomy labels only; no submit-disabled logic, validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Guardrail summarySection-level copy
  • Submit-disabled logic stays unchanged.
  • Validation stays unchanged.
  • No backend, query, or mutation changes.
  • No provider, search, fetch, or crawling.
  • No source-ledger behavior changes.
  • No automatic moderation.
  • No publishing.

See guardrail summary above for unchanged submit, validation, backend, provider, source-ledger, moderation, and publishing behavior.

Step 1 · Prepare evidenceGuidance group
Before you sign inCompact signed-out summary

What you can do now: review the quality checklist and prepare a source URL, relevance context, and relationship disclosure.

What requires sign-in: queue evidence for human moderation and see owner-scoped status or clarification follow-up.

Signed-out summary only; no backend, query, mutation, provider, source-ledger, automatic moderation, or publish behavior changes.

Evidence quality checklistStatic guidance only

Strong evidence usually includes:

  • Primary source or original document
  • Clear relevance to this topic
  • Relationship or conflict disclosure
  • Enough context for an editor to verify

Likely to be rejected:

  • Vague claims
  • Unverifiable screenshots
  • Duplicate links without new context
  • Promotional submissions without disclosure

No AI scoring, provider calls, source-ledger writes, or automatic acceptance/rejection runs from this checklist.

Submission readinessSubmission state: Sign-in required

Sign in before queueing evidence for editor review.

Submitter next step: Prepare a source URL, relevance context, and relationship disclosure before signing in.

No mutation, provider, source-ledger, automatic moderation, or publish action runs from this readiness guidance.

Field helper summaryInput guidance only
  • Title: name the concrete source or evidence packet.
  • URL: use the primary source link editors should inspect.
  • Why it matters: connect the source to this topic's question, claim, or context.
  • Relationship: disclose whether you are a reader, author, vendor, interested party, or anonymous tipster.

Field helper copy only; no validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Completion summaryRequired fields: 0/4
  • Title: incomplete
  • Evidence URL: incomplete
  • Why it matters: incomplete
  • Source relationship: incomplete

Add all required fields before the queue button can enable.

No query/mutation path changes; no provider, source-ledger, automatic moderation, or publish behavior changes in this summary.

Pre-submit confidenceConfidence state: Not ready

Submitter action: resolve sign-in, profile, topic, field, or length blockers before queueing.

Manual review expectation: editors still review every queued row before it can affect evidence or dossiers.

Confidence guidance only; no backend, validation, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Field length guidanceLength status: Within limits
  • Title 0/140
  • Evidence URL 0/500
  • Why it matters 0/1000
  • Source relationship 0/160

All field lengths are within the current local limits.

Length guidance only; no validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Step 2 · Fill required fieldsRequired inputs · Fields and previews

Scan tip: finish the source link first, then add the editor context that explains why it matters.

Source identitySource identity fields · Title · URL · duplicate guidance when relevant

Title and URL stay paired before duplicate context.

Source URL normalization previewNormalized URL preview: Waiting for URL

Type a source URL to preview the duplicate-matching shape.

Preview guidance only; no validation, duplicate blocking, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Editorial contextEditorial context fields · Relevance · relationship disclosure

Relevance and disclosure stay paired before relationship preview.

Relationship disclosure previewEditor-facing disclosure preview: Waiting for disclosure

Choose how you relate to the source so editors can review context before any decision.

Disclosure preview only; no submission payload, validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Field grouping only; no validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Step 3 · Review and follow upManual editor path · Review and follow-up
Editor review pathAfter queueing: manual editor review
  1. Queued for moderation: the evidence row waits for human review.
  2. Editor checks source, relevance, duplicate context, and relationship disclosure: no automated scoring decides the outcome.
  3. Possible clarification request: editors may ask the original submitter for more context.
  4. Accepted or rejected with submitter-visible outcome: final moderation stays manual.

Review path summary only; no submission, validation, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Ownership and follow-upSubmitter-scoped follow-up
  • Status and clarification follow-up are scoped to the signed-in submitter.
  • Public readers do not receive admin controls or private submitter details from this form.
  • Editors review the disclosure context manually before any accepted or rejected outcome.

Ownership summary only; no new queries, tracking, persistence, backend, validation, provider, source-ledger, automatic moderation, or publish behavior changes.

Post-submit expectationsAfter a successful queue attempt
  • The form clears and a success message confirms the row is waiting for human moderation.
  • Your signed-in status tracker can show owner-scoped follow-up after profile and topic readiness.
  • Clarification requests, accepted outcomes, and rejected outcomes stay manual editor decisions.

Post-submit guidance only; no mutation-path, success handling, reset behavior, tracker query, validation, backend, provider, source-ledger, automatic moderation, or publish behavior changes.

Clarification expectationsIf editors need more context
  • They may request clarification from the original signed-in submitter.
  • Respond from the owner-scoped clarification surface after sign-in.
  • Evidence still needs editor resolution before acceptance or rejection.

Clarification guidance only; no clarification-flow, polling, tracker query, mutation-path, validation, backend, provider, source-ledger, automatic moderation, or publish behavior changes.

Duplicate-context expectationsWhen a link looks familiar

A duplicate-looking link can still help if it adds new context, stronger disclosure, or a materially different source package.

Duplicate-context guidance only; no duplicate matching, warning, blocking, mutation-path, validation, backend, provider, source-ledger, automatic moderation, or publish behavior changes.

Source-quality self-checkManual triage only
  • Primary or original: source material, filings, datasets, transcripts, or first-party statements.
  • Corroborating: reputable coverage or analysis that helps verify or contextualize the claim.
  • Weak or context-only: summaries, screenshots, promotional pages, or opinion without clear verification hooks.

Editors still make source-quality judgments manually before any accepted or rejected outcome.

Source-quality guidance only; no scoring, ranking, validation, moderation-logic, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Outcome expectationsAccepted or rejected
  • Accepted means editors found the evidence useful enough to attach to the topic or dossier workflow.
  • Rejected means editors resolved the row without using it.
  • Signed-in submitters may see accepted or rejected outcomes in their owner-scoped status tracker.

Editorial outcome guidance only; no status-tracker query, outcome mutation, validation, backend, provider, source-ledger, automatic moderation, or publish behavior changes.

Final package reviewBefore queueing
  • Re-read source quality: primary/original, corroborating, or weak/context-only.
  • Re-read relevance: why this evidence matters for the topic.
  • Re-read relationship disclosure so editors understand your source relationship.
  • Re-read duplicate context and explain what is new if the link looks familiar.
  • Re-read expected follow-up: owner-scoped status, possible clarification, and manual accepted/rejected outcomes.

Package review guidance only; no validation rule, scoring, ranking, backend, query, mutation-path, provider, source-ledger, automatic moderation, or publish behavior changes.

Manual source credibility checksEditor review only
  • Editors check source identity: who produced or controls the material.
  • Editors check provenance: where the material came from and whether it can be traced.
  • Editors check corroboration against other evidence or context when available.
  • Editors check conflict or disclosure context before relying on the submission.
  • Editors check topic relevance before using the evidence in the workflow.

Source credibility guidance only; no scoring, crawling, provider call, source-ledger mutation, backend verification, validation, query, mutation-path, automatic moderation, or publish behavior changes.

Guidance organization only; see the guardrail summary above for unchanged behavior.

Sign in to submit evidence.

Title 0/140 · URL 0/500 · why 0/1000 · relationship 0/160

No AI extraction, search, crawling, or provider calls run from this form.

Moderation queue examples

No seeded submissions for this topic yet.

Tip line

Submit a tip

Sign in to submit a confidential tip to the editorial team.