Artificial Intelligence Policy
Newsberg tries to find the balance between utilizing the benefits of innovative technologies — including machine learning and AI — while minimizing the ethical and privacy risks involved. As a one-person newsroom, AI tools can be a lifeline for otherwise expensive and time-intensive work. But they must be treated with skepticism, and governed by clear rules that protect this publication’s integrity.
SPJ Code of Ethics
What Is Generative AI?
Generative AI (GenAI) is an artificial intelligence tool capable of generating new content — including text, video, audio, code, and images — in response to user prompts.
Examples of GenAI tools currently used by Newsberg:
|
Otter
AI-powered audio transcription for interviews and recordings. |
Grammarly
AI typing assistant for grammar, spelling, and AP style checks. |
|
Claude
AI chatbot used for ideation, drafting, design, and research assistance. |
SatchelAI
News-specific AI tool suite for editorial workflow support. |
Our Limitations on AI Use
Newsberg will never use GenAI as the sole or primary creator or reviewer of content.
| → | All AI-generated content must be part of a human-led process. Any time GenAI was involved in any part of the process, that end result must be reviewed by a human before publication. |
| → | Human intervention isn’t always possible with real-time AI elements — such as support chatbots, auto-captioning, accessibility tools, or live-stream platforms. In those cases, AI use will be clearly disclosed. At this time, those tools are rarely used. |
How Newsberg Uses GenAI
Ideation
We may use GenAI for it’s copy editing ability, as well as to help brainstorm ideas or wording, or dig deeper into a concept to develop a story idea.
Organization & Communication
We may use GenAI to help draft communications with sources, such as emails. We do not use AI to conduct interviews or report on our behalf. We may use AI tools to transcribe interviews, and all resulting files are stored safely and securely. Sensitive interviews that are conducted off the record are transcribed by hand and not transcribed by AI tools to protect identifying information.
Sourcing & Fact-Checking
We may use AI tools to aid in verifying information. All published content is reported, written, and reviewed by Newsberg humans.
Copy Editing
Newsberg runs human-generated stories through GenAI tools that check for grammar, spelling, and AP style. The original story and the AI copy-edited story are proofread side-by-side and fact-checked after that process to ensure facts, quotes, and details have not been changed.
Labeling AI-Generated Content
All content where any significant element is generated or co-generated with GenAI tools will be labeled as such — in the text body, caption, editor’s note, or another clearly identifiable location.
We will not label GenAI use in purely mechanical processes such as sorting, organizing, or processing existing data; in pre-production ideation; or in non-production applications (e.g., asking an AI tool for a list of social media ideas).
Privacy & Data Security
We will never submit sensitive or protected information through open, proprietary, or shared AI platforms — such as Claude or Otter.ai — where data privacy cannot be controlled. This does not restrict the use of publicly available information in discovery or research through AI tools.
Ethical Standards
At Newsberg, we want to honestly acknowledge the risks and disruptive realities involved in adopting new technologies. We believe AI can increase the efficiency, quality, and effectiveness of human work — and we aim to use it to enhance, rather than replace, human judgment in our operations. As a free, solo operated, local news outlet, we need all the help we can get.
We will also work to reduce and mitigate any harm that may result if technology causes a job or role to become obsolete. That may include providing additional training, redefining roles, expanding responsibilities, and empowering people to be the overseers of technology rather than its casualties.
Acknowledgments
These policies have been heavily inspired by and adapted from the work of many newsrooms and organizations that set the standard for fairness and equity in journalism. We specifically want to thank and attribute fellow LION Publishers member Afrolanews.org, as well as the Institute for Nonprofit News, Chicago Sun-Times, Wisconsin Center for Investigative Journalism, Wisconsin Watch, WCPO Cincinnati, Prison Journalism Project, Tiny News Collective, and Trusting News. This is a fast-moving topic, and only through collaboration can we develop an ethical framework.
Questions?
Any questions about Newsberg’s AI policy can be directed to Founding Editor Branden Andersen at editor@newsberg.org.