AI Usage Policy
We do the analysis ourselves. AI may assist, like merging multiple datasets together or pulling in examples from a hundred different pages, but the actual thought process of what something means, or it means anything at all, relies on our experience and expertise.
All recommendations are ours. If we're recommending it, we've vetted it as being something important to do. The recommendation may include details from AI, like example code written by AI, but we'll have tested and confirmed those details.
All ideas shared are ours. What we share publicly, what we write in audits or other client communication, how we write a recommendation, etc. is from our own thought process. Usually it is our words directly that get shared though we may use AI to find a better way to say something. We may also use AI to help come up with a visual or chart to help clarify a point. We may use AI to brainstorm new ideas, but only as a rough starting point.
We do use AI to organize projects. This includes using AI notetakers during meetings. We are careful with sharing sensitive data, and only do share with client approval. We verify any summaries or suggestions provided by AI against actual files, meeting transcripts and our own common sense.
We will disclose when AI is used. Some projects may require more extensive AI use and be responsible for a more significant portion of the output.
We use multiple models. When AI is used, we do not rely on a single model or a single prompt. We use a variety of AI tools, models and prompts to arrive at the most helpful output.