AI Usage and Disclosure Policy

AI Usage and Disclosure Policy

Draft Notice

This AI Usage and Disclosure Policy is Draft v0.1 for Founder Review and Legal Review. It is Not Attorney Reviewed, Not Board Approved, and Not Published. It is not legal advice and is subject to applicable law and attorney verification.

Purpose

This policy describes proposed boundaries for artificial intelligence use in public website operations, PoliticalPedia civic-information workflows, drafting support, and public disclosures.

Scope

This policy applies to PAC-controlled public content, internal drafting support, public-facing AI-assisted features, and governance review workflows. It does not approve any specific external model provider, dataset, or automated decision system.

Definitions

• AI tool: Software that generates, classifies, summarizes, predicts, or transforms content using automated model-based methods.
• Human review: Review by an authorized person before publication or action.
• Civic information: Public information about government, elections, districts, officials, records, or public processes.

Policy Statements

AI may be used as drafting, classification, research-assistance, or workflow-support technology only when human accountability remains clear. AI output must not be treated as authoritative without review.

PoliticalPedia public responses should preserve source boundaries, disclose uncertainty, avoid legal advice, avoid personalized political persuasion, and avoid unsupported inferences about ideology, motives, or voting behavior.

AI must not be used to fabricate citations, invent missing records, impersonate people, make final endorsement decisions, or publish adopted governance content without approved review.

Responsibilities

The Founder or designated owner approves public AI uses. Technical owners document AI-assisted workflows. Content reviewers verify public claims, citations, and disclosures before publication.

User Rights and Responsibilities

Users should understand that AI-assisted public features may have limitations. Users may report suspected errors, missing sources, harmful output, or disclosure concerns.

Prohibited Conduct

Prohibited uses include deceptive synthetic media, undisclosed automated impersonation, unsupported legal conclusions, individual voter targeting, private-data inference, and automated publication of unreviewed governance decisions.

Review and Enforcement

AI uses should be reviewed before launch and after material changes. Incidents should be documented and escalated. Public corrections should follow the Corrections Policy.

Corrections or Appeals Process

Users may request correction or review of AI-assisted public output. Requests should identify the output, date, source issue, and requested correction.

Records and Retention

AI workflow records, model/provider decisions, review notes, and incident records should be retained according to records policy.

Related Documents

• PoliticalPedia Methodology
• Corrections Policy
• Privacy Policy
• Information Governance Policy
• Responsible AI Governance Framework

Version and Status

• Version: Draft v0.1
• Status: Founder Review Required; Legal Review Required
• Owner: Founder / AI Governance Owner
• Review Cycle: At least annually and before material AI feature changes
• Legal Review Flag: Required; Not Attorney Reviewed
• Board Approval Status: Not Board Approved
• Publication Status: Not Published

Revision History