AI Statement
Last updated June 6, 2026
The conversation around AI has gotten weirdly binary. You’re either all-in or you’re out, and I don’t find that I’m firmly in opposition or in support.
I’m officially in the middle camp.
I’m aware of the real costs, theenvironmental impact, the data center boom, the energy requirements. I do take those implications seriously. I live in an area where data centers are popping up left and right. But I also think refusing to engage with AI is its own kind of problem, especially for someone who does this work. Staying current in SEO and content strategy means using the tools, understanding what they actually do, and forming real opinions from the inside. Willful ignorance by keeping your head in the sand isn’t an ethical position that aligns with my beliefs.
I decided to publish this policy because I want to be clear about where I actually stand and because the people who hire me deserve to know how I work.
My Core Commitments
Transparency. I won’t ever pretend AI-assisted work was done entirely by hand. The strategy, analysis, and recommendations are mine. AI is a research, drafting, and pattern-matching tool that helps me move faster and stay thorough. If you inquire, I’ll tell you were and when it was involved.
Human judgment first. Every audit, deliverable, and recommendation is built on my own analysis, written in my voice, and reviewed by me before it reaches you. AI never gets the first word, nor the final word, on your business.
Your data stays protected. I do not, and will not, feed your proprietary information, login credentials, or sensitive business data into public AI tools or anywhere it could end up in third-party training sets. Hard boundary. If you have specific data handling requirements beyond this, we’ll go over them during onboarding.
No shortcuts that compromise quality. I don’t use AI to fake things like expertise and authority or skip the practitioner-level thinking that’s the reason you’d hire me. If something requires my judgment, my judgment is what you get.
No misleading work. I don’t use AI to manipulate, mislead, or rush anyone into decisions. That includes not generating fake examples, fabricated data, or case studies that didn’t happen.
Bias and accuracy checks. AI outputs get fact-checked, voice-checked, and reviewed for the shallow thinking AI tends to produce when left unsupervised.
Tools I Use
My primary AI tool is Anthropic’s Claude. I pull in other tools when warranted, but Claude is where the bulk of my AI-assisted work happens. I make these choices based on data handling, output quality, and how well each tool does a specific job.
Worth noting: most of the SEO platforms I use day to day, including Ahrefs and Semrush, have AI features built directly into them. That trend is not slowing down. My goal isn’t to avoid AI. It’s to use it responsibly in a way that is optimized for minimal energy consumption by optimizing my interactions as much as possible and genuinely benefits my clients and supports my work. That means knowing which features to use and applying my own judgment to the difference.
Where I Use AI
Pulling and processing data. I like to use AI to aggregate data for me. This is a major time saver. I also use it for pattern-spotting across data points if I can’t find any, summarizing long technical audit content, structuring data in a readable format.
Copywriting. Occasionally I use AI to generate copy for this website, sometimes in emails, and occasionally for social media. What I will do is use it to draft a frame, and then I will rewrite it in my own words. I find it helpful to have a first draft iterated in front of me so that I can eliminate the blank page syndrome.
Drafting from my own analysis. Once I’ve done the strategic work, I use AI to help compile findings into readable, organized deliverables. The way my brain transposes information is NOT generally structured in a way that is helpful for you to read because it often comes out of my brain as a stream of conscious thoughts in a spider web of relational ideas. Typically I record myself walking through your website, then take my own transcript and notes and put them into a format that’s actually easy for you to digest.
Auditing my own work. Running deliverables against my own frameworks to catch anything I might have missed. I am neurodivergent and I am a parent. When you pay me, I want to triple check that I have not missed any of your deliverable outputs.
Project and business operations support. Keeping notes, tasks, and project documentation organized so nothing gets lost.
Occasional ideation. Less often than people assume. Most of my thinking happens in my own head, in conversation with colleagues, or just in the course of doing the work.
Where I Don’t Use AI
Strategy, analysis, and judgment are mine. Always.
Final deliverables do not leave my hands without a line-by-line review and edit.
Most client communication. When you hear from me, you’re hearing from me. I do have some automations and sequences set up with pre-engineered emails that I wrote ahead of time to common questions or inquiries so that you’re not hindered by my capacity to respond.
Image generation. I do not generate images or graphics using AI. This includes Pinterest graphics.
A Note on AI and SEO
The SEO industry is mid-shift, both in how search engines work and in how the work gets done. Agentic search, agentic business support, and the broader technology landscape are changing fast. My position is straightforward: AI is a powerful tool when used by someone who knows what they’re doing, and a liability when it’s used to replace the thinking. I prefer the former. I keep up on AI developments in SEO, but I will always err on the side of human judgment.
If You Think I’ve Crossed a Line
If you believe I’ve violated something I’ve committed to here, I would like to know. You are encouraged to email me info@julieklemens.com with the example in question. I will read and respond to every message personally. I’m a solopreneur, not a corporation, so there’s no formal process. What there is, is little ol’ me, promising to take it seriously.
When This Policy Gets Updated
I’ll update this page if and when my approach meaningfully changes. It is reviewed regularly. The date at the top reflects the most recent revision.
Off-cycle updates will happen if:
Someone raises a concern that reveals a gap in this policy
A tool I use gets acquired by a company with different data handling practices
A credible report surfaces that a tool I use has been training on client data without disclosure
Significant regulatory changes pass that affect how I’m required to disclose AI use
I add a meaningfully different AI use case to my workflow
I drop or add a tool in a way my audience would care about
Okay, back to our regularly scheduled programming!