Our newsroom and readers are increasingly exposed to artificial intelligence or AI, from new digital assistants, to smart homes, to search engine summaries.
Although generative AI has the potential to improve newsgathering and accessibility, the inaccuracies and biases in many current AI tools pose a major risk to journalistic credibility.
We’ve heard important questions from readers and our staff around the ethics, accuracy and environmental impacts of AI. We’re taking these questions seriously as some of them tug at the core of our very existence as a news organization (you can read our reporting about AI here).
Our approach to AI, like everything we do at The Narwhal, is rooted in preserving the trust we’ve built with our audiences.
We want to ensure we have guidelines that align with our values to help us navigate possible AI tools and uses. We’ve developed an AI working group to lay out some organizational guide-posts — not to close ourselves off to new ideas or new technologies, but to make sure we continue to produce accurate, responsible and ethical journalism.
Our working group and guidelines are meant to help us navigate the many diverse and divergent consequences of using and not using AI. We want to acknowledge the uncertainty of this technology, centre humanity and be open to evolution.
AI means different things to different people. Some may call anything a computer does AI. Others may use it to refer to chatbots or large language models.
Our current focus is on generative AI: a type of artificial intelligence that creates new content, such as text, images or media, based on inputs (e.g. “analyze this spreadsheet and provide a list of key insights”) and the data it was trained on.
As we proceed, the following principles will guide our work.
Our AI guidelines
Humanity first
We do not use AI to write articles, or to create photos, videos or illustrations. Reporting will always be led by our journalists, in line with our code of ethics.
No manipulation of people or events in photos or videos
We do not use AI to recreate faces, change expressions or add or remove anything from scenes.
Accuracy and verification
All information generated or tasks supported by AI requires human verification. For example, if we use an AI tool to help us transcribe the audio of an interview, we will always listen back and check to make sure any quotes are accurate. When publishing AI-voiced audio versions of written articles, we will strive to ensure correct pronunciation of interview subjects’ names, other key references and non-English words, with Indigenous words a priority. Everything we publish will live up to our long-time standards of verification, which includes monitoring and addressing biases, ensuring fairness and equity in our journalism.
Transparency
If we use generative AI in the reporting process of a published story, other than transcribing interviews, we will document and describe the tools with specificity in a way that explains and educates. For example, if we use an AI tool to scan and analyze documents, we will fact-check findings against primary source materials and share our methodology with readers. Any uses of AI must abide by The Narwhal’s privacy policy and be raised by staff and freelance reporters or photojournalists with their editors ahead of time.
Accountability
The trust we’ve built with our audience requires us to stand behind the facts and details in every story published by The Narwhal. Any errors or inaccuracies resulting from the use of AI tools will be transparently addressed and corrected, as per our existing corrections policy.
Curiosity and caution
With our values as our foundation, we will approach exploration and experimentation with curiosity, caution and clarity. We will always stop to ask ourselves a series of questions to understand: how could this help or hinder our work?
Evolving practices and guidelines
The Narwhal’s internal AI working group, made up of staff from across the organization, will regularly evaluate and update our standards to ensure uses and tools are values-aligned, equitable and minimize bias. The Narwhal’s AI working group regularly seeks out new research, updates the wider team, stays apprised of concerns and risks, looks to organizations like Trusting News and Poynter for guidance and engages with the wider journalism community on best practices.
Your feedback
This process is evolving as we learn more. Please reach out to us at [email protected] with questions or feedback. If you have a moment, fill in this survey about AI in the newsroom.
These guidelines and the AI survey were created using Trusting News’ AI kit, Poynter and input from The Narwhal’s AI working group. We borrowed language from a policy developed by Kelly McBride, Alex Mahadevan and Tony Elkins, based on input from more than 75 news leaders during AI, Ethics and Journalism summits in 2024 and 2025. We also used language from Poynter’s template for newsroom generative AI policy.
This page was last updated on Feb. 26, 2026.
