Our approach to artificial intelligence.

AI, and especially generational AI, is transforming marketing and communications, bringing huge potential gains in efficiency and effectiveness while also creating new risks. It’s having a big impact in how we work too, and we want you to know how we’ll use it and won’t use it. Of course, things are changing rapidly, but this is our current state of practices as of 2025.

How we’re using AI today.

  • We use image tools like Midjourney to help visually convey the stories we’re telling. We can do this in ways that align with our client’s brand, create something truly unique, and save money on stock photography or photo shoots. It takes prompt engineering skill to do this well.

  • We use tools like ChatGPT to get smarter quickly. We’re constantly learning about new and often complex topics, from how to use LIDAR to map the ocean floor to how to use digital twins to improve factory floor performance. ChatGPT can be great at helping us quickly grasp the basics and point tor resources to go deeper. That lets us spend more of your budget on ideas and execution.

  • We use GenAI tools to get past the blank page. While these platforms aren’t necessarily overly creative, they get your brain going and help map the topic terrains we can explain. For example, we explored countless AI-generated names for our agency but ultimately the one we came up, Narratio, was our own idea.

  • We use AI to optimize. That can be to recommend keywords, shorten text to fit available space, or simply proof to improve readability. However, humans choose whether to accept any of these recommendations.

  • We use AI to focus on you. We love using tools like Otter to transcribe and summarize meetings so we can really focus on the conversation.

How we’re NOT using AI today.

  • We are not outsourcing article writing to AI tools. Many media outlets expressly forbid this, and at any rate, in our experience, they are not particularly authentic, interesting, or accurate anyway.

  • We are not using photo-realistic AI-generated images for press materials, even if disclosed. Journalists expect images to come from the real world and not be fictions. See above re: article writing.

  • We are not interested in disrespecting artists and creators by stealing their style. For example, we will never ask a visual tool to create an image in the style of an artist’s existing, or a text tool to write something in the style of a well-known author or published piece.

  • We set boundaries on how much we trust AI tools. Yes, they say they will keep your information confidential, but our confidence in their commitments is limited.

  • We do not outsource accountability to AI. If AI gets facts wrong, it’s our job to catch it. If it puts something in an image that shouldn’t be there, it’s on us to fix it. We’ll never blame the AI.