RANGE Artificial Intelligence (AI) policy

Last revision: September 29, 2026

Artificial Intelligence is seemingly everywhere — whether we like it or not. As a worker-owned newsroom and press for the people, RANGE Media has obvious concerns about AI and its negative impacts on our communities, livelihoods and environment. Yet AI, specifically machine learning, has been around for decades and has been long-integrated into tools journalists rely on in our work everyday, such as transcription software and spell checkers. 

We recognize that AI-enabled technology can support our staff in the work they do and help us expand RANGE’s service to our community, but we cannot let it undermine the trust our readers have placed in us and we don’t want to abdicate our responsibility as reporters to machines. As AI becomes more pervasive, it’s crucial to clearly define our own use of the technology and push for transparency around the disclosure of its use in our wider society. 

Guiding Principles

AI, particularly GenAI, is a rapidly evolving technology with new tools emerging everyday and new features thrust upon us at the will of tech companies. No one (even, it seems, the tech overlords themselves) knows exactly where this technology is headed. So, below we have outlined some of our hard-and-fast rules for our use of AI below, but we cannot account for every possible use case. Here are our guiding principles:

• The Sanctity of the Written Word: We are the writers here. We will never use AI to write content or to communicate with sources. Other professions may have different standards, but reporters are trained with a foundational rule: If it’s under your name, and it’s not your words or your editors’, it better be in quote marks. Otherwise, it’s plagiarism. We don’t do plagiarism.
• Transparency: Disclosure of significant use of GenAI in our content should be specific, including the tool used and to what extent. (However, the disclosure of basic production tools such as Otter and Canva is not necessary.) 
• Accuracy and Human Oversight: Verification of information is part of our job. We may use AI at some point within our research processes, but we will always read the primary sources and look beyond what is served to us. We may use AI to transcribe interviews, but we will always listen to the section being quoted to ensure our quotes are accurate. We may use AI to extract or clean data, but we will always validate that data with the source. We may use AI tools to help with Spanish-language translations, but the final say on accuracy will always sit with a native Spanish speaker. 
• Privacy and Security: We will protect our audience’s data in accordance with our newsroom’s privacy policy. Tech companies change their privacy policies often, so we don't rely on their promises. We treat anything we upload to an outside service as something that could be stored, read or used to train AI, and we decide what to upload with that in mind. 
• Learn to Fish: In cases where GenAI is assisting in an operational task that can be reasonably learned, such as basic programming, we should try to learn from the process rather than have it just do the task for us. “Teach a man to fish” and all. 

For editorial and content creation

WE MAY:

• use AI in the same way reporters have used Google and Wikipedia for decades: as a tool for tracking down primary sources, news articles, data troves or other sources of information
• use AI in the same way that reporters use spellcheckers and basic grammar checkers: to flag potential errors. We will consider each suggestion individually, and not simply copy-and-paste something the AI spits out
• use AI to transcribe interviews. But we will always verify the transcription by listening back
• use AI to clean and organize data, either by using a ready-made tool or by helping us write spreadsheet formulas. We will always be clear about how we used the tool and its limitations, and we will analyze the data and draw conclusions using our own human brains. 
• use editing tools embedded in software that utilize assistive AI. For example, Canva’s background remover tool has helped us from RANGE’s earliest days to make our fun photo illustrations with stock photos and Descript’s filler word removal tool saves us hours each week by removing excessive um’s and uh’s in our podcasts 
• use Google Translate and Spanish grammar tools to help with translations
• use AI to compare large documents to quickly identify differences. For example, contrasting multiple versions of proposed legislation being considered
• use GenAI for accessibility features, such as drafting alt text descriptions, with human review and approval 

WE WILL NEVER:

• require our team members to use AI
• use AI to generate written content, graphics, photos, podcasts or videos. This includes using AI to draft articles and “experiment” with sentence structure and wording 
• use AI as the sole editor or reviewer of a piece of content
• input sensitive, confidential source information into large language models 
• accept columns from community members or or articles from freelancers with AI-generated writing or significant GenAI editing

For operations   

WE MAY:

• use AI to assist in website debugging and minor web development, provided we do not enable an AI agent to make changes for us 
• use AI to assist in building internal tools (such as an editorial calendar Slack alert widget that Val is pretty proud of)  
• use AI to assist in complicated budgeting and projection work (such as a growth-scenario planner Luke is pretty proud of)
• use AI to assist in finding and applying for grants, provided final editing and sign-off comes from a human
• use AI to help draft internal process documentation
• use AI-enabled productivity and accessibility tools, provided they do not violate our privacy policy 

WE WILL NEVER:

• base the integrity of our website development on AI models
• use AI to gather private user data without consent
• use AI to generate and implement code without knowing its effects
• implement an annoying AI chatbot on our website

As a worker-owned newsroom, this policy was developed communally. Some members of staff would rather chop off their left hand than use AI to write literally anything, while others see it as another tool to fight the good fight, and even as an accessibility aid. These are the baselines we could all agree on, and no one will ever be forced to engage with AI more than they are comfortable. Robots will never truly understand this thing humans call love. 

Special thanks

Rather than reinventing the wheel (or having AI generate it for us), this policy has been partly adapted from the policies and work of newsrooms and organizations that have set the standard for our industry. We specifically want to thank Trusting News, Poynter Institute, AfroLA and Nucleo.