AI in MTN newsrooms: practical workflows writers can use without losing the plot
AI is in the newsroom, but it isn't writing MTN's stories. Reporters and producers still do the legwork: research, interviews, scripting, and fact-checking. AI steps in for speed and formatting - with human oversight at every stage.
Where AI actually helps
MTN teams have access to an internal assistant called Engine Room. Some use it, some don't. The point is optional efficiency, not replacement.
- Document triage: Anchor Andrea Lutz uses AI to scan hundreds of pages of public records, surface key sections, and flag page numbers for quick verification.
- Format conversion: Multi-media journalist Isabel Spartz writes a TV script, then uses AI to transform it into standard AP style for the web in seconds - same reporting, different format.
- Transcription: In their editing software, interviews are transcribed fast, so producers can focus on structure and storytelling.
The guardrails that keep it honest
"You have to check the AI because it's not always accurate," Spartz said. Lutz double-checks every summary and asks for page citations to confirm the source. Producer Cody Boyer adds a simple rule: always put another set of human eyes on anything AI touches.
That's supported by training sessions on the tool and a culture that treats AI like spellcheck with extra steps - useful, but never the final say.
A cautionary example: fake quake photo
During an interview with Boyer, the team found a viral social post claiming to show Great Falls after a recent earthquake. It came from a Facebook account built on AI-generated images and had 34,000 followers. It looked convincing. It wasn't real.
This is why verification still matters. If you publish, you're responsible. If you're a writer, consider keeping a quick reference to ethics best practices like the SPJ Code of Ethics.
What this means for writers
Borrow the newsroom playbook. Keep your voice. Let AI speed up the grunt work.
- Use AI to scan long PDFs and pull highlights - require page numbers for each claim.
- Convert drafts across formats (script to article, outline to newsletter) without changing your core message.
- Transcribe interviews and calls, then verify quotes against the audio before publishing.
- Fact-check everything AI outputs. If it can't show the source, don't use it.
- Build a two-minute pre-publish checklist: sources cited, quotes verified, links tested, claims supported.
- Keep a second reader in the loop for sensitive topics or data-heavy sections.
If you're exploring tools and workflows for writing, this curated list is a helpful starting point: AI tools for copywriting.
What stays human
MTN's stance is clear: reporting, interviews, editorial judgment, and ethics are human work. AI helps with speed and structure - nothing more. As Lutz put it, the fundamentals don't change.
Transparency in practice
MTN discloses when AI helps convert on-air reporting to a web article and states that editors verify all versions for fairness and accuracy. That's a simple, repeatable model: use the tool, check the work, tell the reader.
Bottom line for writers: Keep your standards high, your sources verified, and your workflow efficient. Let AI handle formatting and transcripts. Keep the judgment - and the voice - yours.
Your membership also unlocks:
