The Foresighter — Policy on the Use of Artificial Intelligence

AI assists our journalism; it does not author it. Human editors are accountable for every word, number, and image we publish. We disclose synthetic visuals, verify AI-assisted research against primary sources, protect data, and correct errors plainly.

This policy governs how staff, contributors, and contractors at The Foresighter may use AI systems (including LLMs; transcription, translation, image, and audio tools) across reporting, editing, design, production, audience, and business operations.

First Principles

• Human judgment first. Editorial judgment, analysis, and narrative voice are human responsibilities.
• Independence. AI tools support our work but never set editorial stances or coverage priorities.
• Transparency. We disclose material AI involvement in visuals and any other use where a reasonable reader might be misled.
• Security & privacy. We protect confidential sources, drafts, and sensitive data when using any tool.
• Accountability. Editors are accountable for AI-assisted outputs and must verify facts against original sources.

   

Permitted (Assistive) Uses

AI may be used to:

• Research triage: surface leads, cluster documents, generate reading lists, and create search strings.
• Editing aids: grammar, style suggestions, clarity passes, headline and dek variants, and caption drafts.

   

Prohibited Uses

• No AI-written articles or bylined pieces. Final prose is human-written and edited.
• No synthetic quotes or invented sourcing. Ever.
• No undisclosed synthetic news imagery (including photorealistic depictions of real events/people that could mislead).
• No use of AI to decide editorial positions or to fabricate audience sentiment.
• No uploading of protected/confidential materials to tools without approved safeguards.

   

Disclosure & Credits

• Text. We do not label routine AI assistance (e.g., grammar) per piece; this policy serves as disclosure.

   

Verification & Editorial Control

• Source of truth. Facts derived with AI assistance must be confirmed against primary sources (documents, datasets, transcripts, filings).
• Hallucination guard. If a tool provides a claim without a verifiable source, it is treated as unsubstantiated until independently verified.

Visual Standards & Synthetic Media

• News photography/video. Only standard, non-deceptive edits (crop, exposure, color balance). No removal or addition of elements.
• Conceptual art. AI-assisted illustrations may be used in non-news, clearly conceptual contexts with disclosure.
• Deepfake bans. No lifelike depictions of real individuals in fabricated contexts.

   

Freelancers & Contributors

• Same rules apply. Freelancers may use AI only for permitted assistive tasks and must disclose any material AI involvement in visuals.
• Warranties. Contracts require originality, IP compliance, and adherence to this policy.

   

Governance & Review

• Ownership. The Editor-in-Chief and Standards Editor own this policy; the CTO/Tech Lead owns the tool registry and security controls.
• Cadence. Reviewed at least twice a year or upon major model/tool changes; a changelog is maintained on our “How We Work” page.

   

Contact: standards@theforesighter.org (Editorial Standards)

This policy is a living document. If you see a better, safer way to use AI in service of rigorous, optimistic journalism, tell us—and we’ll improve it.