Policy
Last updated: April 9, 2026

AI Transparency & Ethics

Artificial intelligence is reshaping journalism. We believe readers deserve to know exactly how — and how not — AI is used in producing the content they read. This policy governs all AI tool use at NonProfit Media.

Our core commitment: AI is a reporting tool, not a reporter. No article, investigation, or analysis published by NonProfit Media is written by AI. Every word is written and approved by human journalists.

Core Principles

Disclosure

Any content that used AI tools in its production is labeled with an "AI-assisted" disclosure. Readers have the right to know when and how AI was involved in creating what they read.

Human Primacy

Every piece of content published by NonProfit Media is written, edited, and approved by human journalists. AI is a tool, not an author. Final editorial authority rests with our human staff.

Accountability

Errors introduced or amplified by AI tools are subject to the same corrections process as any other error. The use of AI does not diminish our accountability to readers.

Bias Awareness

We acknowledge that AI systems can reflect and amplify biases present in their training data. We actively audit AI-assisted work for bias and do not use AI tools for coverage decisions about marginalized communities without additional human review.

Source Integrity

AI tools are never used to generate, fabricate, or embellish source material. All quotes, documents, and source claims must be independently verified by human journalists before publication.

Vendor Transparency

We disclose the categories of AI tools we use. We do not use AI systems that cannot explain their outputs, that have opaque training data, or that are operated by entities with conflicts of interest in our coverage areas.

What AI Can and Cannot Do at NonProfit Media

Permitted Uses (With Human Oversight)

  • Writing or drafting news articles, investigations, or analysis (human-edited and fact-checked)
  • Generating quotes attributed to real people (verified with original sources)
  • Creating synthetic images, audio, or video of real individuals (clearly labeled as AI-generated)
  • Making editorial judgments about newsworthiness (final decision by human editors)
  • Transcription of audio/video interviews (human-verified)
  • Translation of foreign-language documents (human-reviewed)
  • Data analysis and pattern identification (human-interpreted)
  • Search and retrieval across document archives
  • Grammar and clarity editing of human-written copy
  • Generating structured summaries of public records (labeled)
  • SEO metadata and headline suggestions (human-approved)
  • Fact verification assistance (all AI output independently verified by journalists)
  • Source research and interview preparation (journalists conduct actual interviews)

AI Tools We Currently Use

We disclose all AI tools used in our editorial workflow. This list is updated when tools are added or removed.

CategoryToolUseHuman Review
TranscriptionWhisper (OpenAI, open-source)Audio/video interview transcriptionYes — all transcripts reviewed before use
TranslationDeepL ProForeign-language document translationYes — bilingual staff review for accuracy
Document AnalysisInternal tooling (LLM-assisted)Pattern identification in large document setsYes — all findings independently verified
Editing AssistanceGrammarly BusinessGrammar, clarity, and style suggestionsYes — all suggestions reviewed by author
Search & RetrievalInternal vector searchArchive search and related-article suggestionsN/A — no content generation

Deepfakes & Synthetic Media Policy

NonProfit Media will never publish synthetic images, audio, or video of real people without explicit disclosure that the content is AI-generated and without a clear editorial justification (e.g., illustrating a story about deepfakes themselves).

When we cover deepfakes and synthetic media as a news topic, we follow these protocols:

  • Synthetic media examples are clearly labeled "AI-GENERATED EXAMPLE" in all contexts
  • We do not republish deepfakes of real people without editorial necessity and legal review
  • We consult with digital forensics experts before reporting on the authenticity of disputed media
  • We follow the Coalition for Content Provenance and Authenticity (C2PA) standards for content verification

Accountability & Feedback

We review our AI use policies annually and update them as the technology and journalism landscape evolves. We welcome feedback from readers, researchers, and journalism ethics organizations.

If you believe AI was used inappropriately in a piece of our content, please contact our Standards Editor at [email protected] with the article URL and your concern. We will investigate and respond within 10 business days.