AI Use Policy
Last updated August 2026
Applies to public content, marketing, website copy, administrative work, and research. Clinical documentation and telehealth are covered separately in the Telehealth and AI-Assisted Documentation consent.
TL;DR
I write my own content. My brain runs fast and full, and I write to get useful things out of it and to other people in hopes that it is helpful. But often my very full, fast-moving brain means I'm not a great writer. I'm long-winded, grammar is terrible, I jump from topic to topic, and my writing often reflects the chaotic inside narrative, only without the insight into how I got from A to B. Without the assistance of AI in my writing, I most likely wouldn't write anything I felt I could put into the world. I use AI mostly to edit and help get my points across. I write, AI edits, and I finalize.
I also hold real concerns about what AI costs, from the environment and the people whose labor and creative work built it to bias, concentrated power, and what it does to how we learn and think. I haven't found a clean resolution to those concerns, and I do know there is much we all have to learn on the topic. So I make specific choices inside this complexity. I choose tools by how they govern themselves and handle data, I keep and use my books and use reputable sources, and I advocate for caution and regulations in adopting this technology. The same ethics and care that guide my work as a licensed counselor, to stay informed, think things through, and act from my values, apply here too.
AI use policy
I write my own blog posts. There's rarely a limit to the information racing through my brain or how much of it I want to share. Especially if it might help even one person. I started my website blog because marketing told me I should. Once I started I realized there are some pluses beyond marketing. Creating a blog pushes me to think deeply about a topic, including perspectives that aren't mine, sharpen what I believe, and learn a topic properly (I read the academic research, other people's blogs, and go down the occasional four-hour rabbit hole; hello, ADHD brain). What starts as one post often turns into a series, or into teaching material for clients or other therapists.
Honestly, though, I'm a shit writer. I'm long-winded, I don't transition worth a damn, and I get so caught in my own racing thoughts that I lose the point I wanted to make in the beginning. I also want my writing to be found, so it reaches more than the people who already know me, and I don't know how to do that alone. My business isn't yet at the point where I can pay someone who does. One day, I hope it is.
This is where AI comes in. I use it to brainstorm when a topic might work better as a series than a single post. And I use it to edit, tighten sentences, check structure, and catch where I've buried the point. The ideas, the clinical perspective, and the final words in every sentence are mine. I write, AI edits, and I finalize. Then I hope someone gets something from it.
Writing this statement has been on my to-do list for a while. But as with many tasks in my life, I sometimes need an extra push. The new watermarks for AI content arriving on the scene (hello, sensible regulation) are what finally pushed me to write it up, so my transparency stays ahead of it. What follows is how I think about and use AI throughout my work, including the complexity of the topic and the concerns I hold.
What this covers
This policy is about how I use AI in the non-clinical parts of my work like blog posts, marketing, website copy, and administrative tasks. My policy regarding AI use and a broader statement on technology for sensitive client-facing work is documented in my informed consent and practice policies provided to clients and colleagues. This policy is one piece of that larger system.
The standard I hold myself to
I am a licensed professional counselor. The obligations that come with that, to stay informed, think decisions through rather than defaulting to convenience, and act from ethics and values, extend to my use of technology. I approach AI the way I approach any tool that touches my work and people.
How I use AI
Here are many of the ways I use AI at this time.
Brain dumping thoughts to help organize, prioritize, and create a plan.
Remember where I left off.
Thinking through structure, when I know what I want to say and need help getting the words out of my head.
Editing to tighten sentences, cut repetition, and catch where I've buried the point or drifted from my own voice.
Drafting functional material like templates and administrative documents that I then adapt.
Adapting material for new scenarios.
Creating some handouts, marketing materials, and PowerPoint slides.
Checking my own reasoning, the way I would talk something through with a sharp colleague, keeping what is useful and dropping what is not.
How I do not use AI
I do not use, nor trust, AI to provide me with well-rounded information, create content without first heavily drafting and later reviewing it myself, or provide me with information without providing sources that I then review.
I do not take as fact anything generated with AI and I question anything that doesn't make sense and any possible hallucinations.
I do not let AI make clinical decisions or generate clinical content. My clinical consent covers that.
I do not enter client information, or any identifying detail, into a tool that lacks a Business Associate Agreement.
I make every effort to not allow AI to sway or inform my opinions, biases, or judgments.
I do not allow AI to impact my emotional well-being or lack of.
Authorship, review, and watermarks
I write my content. When AI touches a post, it edits what I already wrote. I write, AI edits, I finalize. The ideas, the clinical perspective, and the final version are mine, and I am accountable for what I publish no matter what helped me get there.
On watermarks, as of 2026, text edited by some AI tools carries an invisible, machine-readable mark showing AI was involved at some point. That mark does not identify who wrote what, and it does not track how much content a human changed. My process already keeps the writing and the final edit in my own hands. However, despite this, I have no way of knowing at this time if my writing is edited enough to not appear as though it was authored by AI.
The risks and potential benefits
There is more to say about AI's costs and its effects on people, learning, and society than I will pretend to know or fully understand. This is true of a lot of things in life, including most technology. I don't throw my hands up in the air over this, but rather I seek to learn where I can and remain humble about what I do not know.
I understand there are real costs, risks, and potential benefits associated with AI. Very similar to how there are risks and benefits to therapy. Biases and assumptions, concentration of power and knowledge, exploitation, monetization of learning, and risks to critical thinking are present. Each of these areas can increase safety risks for us all.
Furthermore, the environmental cost is deep. These projects take land, water, and peace, often from communities with the least power to refuse. Within my own state of Tennessee, we are fighting to protect land and wildlife rather than profits and technological advancement.
I also acknowledge that this technology itself isn't all bad. There are wide opportunities for increased knowledge and research, improvement in health outcomes, ability to think with greater complexity through societal problems and bring about solutions, and AI-generated tools that can increase accessibility for neurodivergent and disabled individuals, myself included.
So, in addition to acknowledging the risks and benefits of AI, I also try to advocate for reduction in harm around its use. I stay informed, I show up to the discussions, I educate others, I vote, and I use my voice, including here. I am for sensible regulation and rules that help us steer this toward good, because it can be good.
I live and work inside a system I did not design, and thoughtful use is the best option in front of me. So I make values-based choices where I can, and I accept complexity where I cannot. I choose tools based on how the company governs itself, how it handles privacy, and what it commits to doing with data, and I have declined tools that failed those tests. I do not simply trust what vendors say or even organizational ethics statements, because trust is earned and organizations shift their values. I research what I am told and I strive to thoughtfully keep up with relevant changes in the tools I use.
Review
I review this policy at least once a year, and any time I add, change, or stop using a tool. Next scheduled review August 2027.
Contact Us
If you have any questions about this AI Policy or how we handle your data, contact us at:
📧 Email:Hello@KestrelPsychotherapy.com
📍 Business Name: Two Birds Development Company, LLC
🌐 Websites: KestrelPsychotherapy.com | TwoBirdsDevelopment.com | RandiWilliamsLPC.com