# BoiseDev's AI policy: humans write the news, AI helps find the facts

> BoiseDev and Valley Lookout say humans write the news-reporters keep the voice, context, and accountability. AI can help with research and production, with labels and disclosures.

- Type: AI news
- Page: https://completeaitraining.com/news/boisedevs-ai-policy-humans-write-the-news-ai-helps-find-the/
- Published: 2026-01-14
- Updated: 2026-02-19
- Author: Jeroen Erne
- Jobs: Writers
- Reading time: 3 min

### Humans should write the news: what BoiseDev's AI policy means for writers

AI will touch every corner of publishing. That's not a warning. It's a prompt to get clear on what work stays human-and where machines actually help.

BoiseDev and Valley Lookout put stakes in the ground: reporters write the stories. AI can assist [research](https://completeaitraining.com/tag/research/) and production, but it doesn't write the news. If you're a writer or editor, there's a lot here you can borrow.

#### The core principle

Humans write for humans. That's the non-negotiable. Voice, context, and accountability sit with the reporter, not a model.

#### What BoiseDev will and won't do

- **No AI-written news copy.** Not whole articles, not paragraphs, not "light touches."
- **AI as an assistant only.** Use it to analyze documents, sift databases, transcribe calls or hearings, summarize long records, and spot leads.
- **Clear disclosure.** If AI meaningfully informs reporting, note it in the story.
- **Labeled visuals.** AI-created graphics get explicit labels.
- **Editing tools are fair game.** Spellcheck and grammar aids can flag issues; an editor decides.
- **Policy can evolve.** Changes happen in public, with transparency to readers and staff.
- **Ad exception.** These rules don't apply to advertising materials.

#### Why this matters for writers

- **Trust.** Readers can feel the difference between lived context and stitched text.
- **Attribution.** A reporter earns trust by citing sources and verifying claims. A model can't hold that responsibility.
- **Originality.** Your voice is the product. Outsource it and you compete on volume, not value.

#### Case study: AI as a research assistant

During Rite Aid's bankruptcy, a BoiseDev reporter had puzzle pieces: store signage, a permit trail, and a phone confirmation about a closure. The gap was buried in court filings that bundled many stores.

A chatbot helped locate a key line: the building's owner objected to transferring the site to CVS. That lead was verified against the original filing, then woven with on-the-ground reporting and archival notes. The story was written by a person. The tool surfaced a needle; a human built the narrative.

#### What to avoid

- **Ghostwriting by AI.** Don't let models draft news copy. Even "first passes" leak style and phrasing you didn't earn.
- **Unlabeled use.** If AI materially shaped the reporting, say so.
- **Plagiarism via automation.** Spinning someone else's work through a chatbot is still theft.
- **Ungrounded claims.** Models fabricate. Always verify with documents, calls, and firsthand reporting.

#### Headline pledge: accuracy over clickbait

BoiseDev calls it out directly: headlines should reflect the story. Clever is fine; misleading isn't. For writers, that's a simple rule-match intent, set expectations, and deliver.

#### A practical policy you can adapt

- **Write your non-negotiable.** "Humans write the news. AI assists research and production." Put it in your style guide.
- **Define approved use.** Research synthesis, doc search, transcription, data exploration, and draft outlines are OK. Writing copy is not.
- **Set disclosure rules.** Add a one-line note when AI materially aids reporting or analysis.
- **Label AI visuals.** Keep it obvious for readers.
- **Build a verification loop.** Every AI-sourced lead must be traced to original documents or direct confirmation.
- **Protect your voice.** Keep a house style. Editors enforce tone, structure, and rhythm.
- **Log your steps.** Save sources, prompts, and citations. Make review easy.
- **Review quarterly.** Update the policy in public. Note what changed and why.

#### Context: when this shift started

[ChatGPT](https://completeaitraining.com/tag/chatgpt/)'s public release in late 2022 pushed AI into the newsroom conversation overnight. Many shops leaned into efficiency; some crossed lines with unlabeled or derivative copy. BoiseDev chose clarity and reader trust instead.

[See the original ChatGPT release](https://openai.com/blog/chatgpt?ref=completeaitraining.com)

#### If you're a writer, steal the process-keep the craft

Use AI to speed the grunt work: find filings, summarize transcripts, surface patterns. Then do what only you can do: call, confirm, connect dots, write with judgment, and accept the responsibility that comes with a byline.

That's the line BoiseDev drew. It's a good line for any writer who cares about trust, long-term career value, and work they're proud to sign.

#### Want structured practice with AI (without losing your voice)?

Explore practical courses for your role so you can use tools for research and workflow-not for replacing your writing. Start here: [AI courses by job](https://completeaitraining.com/courses-by-job/).

---

This is the Markdown form of https://completeaitraining.com/news/boisedevs-ai-policy-humans-write-the-news-ai-helps-find-the/. The whole site is mapped at https://completeaitraining.com/llms.txt.
