On this week’s episode of Katy Calls BS, I launched something I’m calling the AI Resistant Pledge. It’s pretty straightforward. No BS. Imagine that.
The AI Resistant Business Pledge
I’ll let AI touch my drafts, not my judgement.
I’ll use it to go faster, not to think for me.
My voice, my face, and my point of view stay mine.
That’s it. That’s the whole thing. No huge certification, no course, no sign-up form. You don’t have to send me anything. It’s something you can take for yourself or not. Add the #AIresistantbusiness if you want.
Here’s what the AI Resistant Pledge is all about:
Each line points to a decision you make in your business so we can be less boring and less mainstream. If you are reading this blog or listening to my podcast, chances are, you like to do things differently. Welcome to the party!
Since the former teacher in me just can’t quit, here’s what it looks like in your day to day operations.
“I’ll let AI touch my drafts, not my judgement.”
AI might produce your first draft of an email, an outline, a social media caption or maybe even a whole campaign for you. The draft though because I get it, sometimes we need help with ideas. It does not get to decide what you think about a client situation, a pricing question, or a piece of advice you’re giving someone who’s paying for YOU or YOUR judgement. You are the final say, you still need to put you in your brand and that does not mean training AI, it means writing in your voice.
“I’ll use it to go faster, not to think for me.”
We are busier than ever. I basically live in my car on the days my children are with me. When they aren’t, I’m driving to their games. Speed is the reason most of us are reaching for AI and why it’s so appealing especially as business owners who wear a million hats. The line gets weird when people start using it as a substitute for their own thinking. The other way is when you just copy and paste the output. That’s how we lost the em dash.
“My voice, my face, and my point of view stay mine.”
I mean, this one is obvious, yeah? I thought so until I saw all the AI brand photos, videos, and heard a video dub of my voice. No thanks! This is a no go for me. As someone who has been in business for almost 8 years, starting in social media marketing, no, we’re not doing this. You lose the trust of your buyer almost instantly if you are not showing your face. I built my brand using video and using AI generated video is not it.
AEO Basics: The Very Basic
Part of using AI as a tool instead of seeing it as another thing we have to learn is understanding how it’s reshaping the way people find you. Answer engine optimization or AEO is the practice of structuring what you publish so AI tools can find it, understand it, and cite it when someone asks a question in ChatGPT or an AI search overview instead of typing it into Google.
I’m not an expert at all but this is where a lot of clients are finding service businesses. It doesn’t require being a technical person and there are experts who can help you. I can send you a few.
- Lead with the answer: Whatever question a piece of content answers, say it simply in the first sentence or two. AI tools pull direct answers, not the buildup to the answer.
- Add a real FAQ section to your key pages: Write out the questions clients ask you in their words with direct answers under each one. This is high-leverage AEO work and most people skip it.
- Keep your facts consistent everywhere: Your pricing, your process, your positioning, try to say it the same way on your website, socials, and show notes if you have a podcast. AI tools cross-reference for consistency. Mixed signals make you less trustworthy and that is the last thing we need!
- Keep things current: AI tools weigh recent information more heavily. A stale page stops getting cited the same way a stale freebie stops converting.
- Show up in real conversations, not just your own content: Answer questions on industry forums, Reddit threads, those get weighted differently than marketing copy because you sound more human.
This is structure work and doesn’t mean you will win the fight with AI. You can still sound generic AF with AI. AEO just gets you found online. People might still decide they don’t like what they find.
Is there such a thing as ethical AI?
I don’t about this. I don’t really think there is a clean answer because there is some sketchy stuff happening lately. The real concerns about AI are pretty well documented. Copyright law hasn’t caught up with models trained on work their creators never agreed to hand over. Training and running these systems at scale carries an environmental risk that I don’t think the public truly understands. Especially if we keep creating ugly AI posters every chance we get.
A small number of companies are making decisions that shape how a huge share of the world finds information and makes choices, with very little outside oversight. It’s not just a talking point, it’s the current state of the industry. And it’s pretty freaking scary to think about it that way. We have incredibly powerful tools that can do more than we ever imagined but at the same time we have incredibly powerful tools that can do more than we ever imagined!
Just this month, September 9, 2026, a researcher named Jacob Coxen, who worked at OpenAI and Anthropic, resigned and said publicly that neither company was “acting responsibly,” warning that systems being built would “soon be superhuman systems that can hack anything, revolutionize any field overnight and acquire real power and resources.” He said the people building the people building this technology “earnestly believe it could kill us by the end of the decade.” Cool. Cool. That’s a strong claim from one person who has direct experience!
Three days later, Anthropic CEO Dario Amodei published a post saying, “We must slow the pace at which we improve the capabilities of AI models.” He made sure to clarify that pacing meant giving companies time to align and safeguard what they build not halting the progress.
Sam Altman from OpenAI responded by pledging to adopt independent evaluators with employee level access to check that work. Elon Musk’s response to all of this was “Dario is right.” At the same time, more than a thousand staffers across major AI companies had already signed a petition asking the US government to support mechanisms that would deliberately slow AI development down.
Truly, WTF. I use AI in my business and I am not bringing all of this up to scare anyone away from it. I bring it up because when the people running the companies building the technology are publicly asking each other to slow down, that’s not a random opinion you can dismiss. It makes you think about how you are using the tools in your day to day operations. Think about how much of your business you hand over to AI.
Ethical AI really depends on how you are using it kind of like ethical business practices. It’s one of the things that will continue to burn out business owners. You have people doing honest work and you have people who will continue to use AI tools to create generic, mainstream websites, logos, courses, digital products, brand photos, social media campaigns, and more and wonder why it looks like everyone else.
New ways to use AI: Pay the experts, don’t replace them
Stop trying to build everything with AI tools and hire the experts who have years of experience. When you hire them, TRUST THEM.
You aren’t rejecting AI and you aren’t making your business better by using AI to do everything for you. Here’s how you can use AI to get more out of your experience with the professionals you hire and still respecting their work.
- Draft your questions before a paid consultation: Before a call with your accountant, brand and website designer, brand photographer, digital product designer (that’s me), run your situation by AI and ask it to generate the specific questions you should be bringing. You walk into that paid hour or discovery call asking sharper questions. This is especially helpful if you aren’t sure what to ask.
- Organize your own documentation before you hand it to an expert: Receipts, contract history, any assets, ideas, etc, have AI sort it out and summarize it before it lands on your bookkeeper’s desk or before you send it to your client. You save time and they get a better summary.
- Get a first read on a contract or vendor proposal, then still send it to a professional: AI can flag missing terms or do research for you. This is great for catching inconsistencies. Maybe you have a vendor event coming up and you want to know how past events went, have AI do some research for you! Ask if this is going to be worth something for you to invest in.
- Screen a client or potential hire for obvious red flags before you spend interview/discovery call time: AI can help you sort faster but can mistakes. There are times when I’ve used this in the client process and I can’t decide if I should move forward, the hint is in the hesitation. The final call was all me but I needed a full picture that I couldn’t see to help me make the final judgement call.
The pattern across all 4 of these is AI does the prep. The person you pay for a service will provide the judgement. That is what they do, it is their zone of genius, let them do it! You get more value out of every dollar you spend when you can trust someone’s expertise instead of trying to replace it with a generic AI version.
Are we trying to have generic AI businesses? I don’t think so.
Where to go from here
If you want the full context of this week’s episode of Katy Calls BS where we talk about the two camps of AI use, the research on what heavy reliance does to your thinking, and how an “AI takeover” happens inside a small business, listen to episode 156, Stop Using AI to Sound Generic.
If you’re not sure where your own business sits on this and you need help adding more judgement in your digital products, not just information, book a Digital Product Infrastructure Audit.
Quick answer: The AI-Resistant Business Pledge is a three-line commitment to let AI touch your drafts and not your judgment, use it for speed rather than thinking, and keep your voice and point of view your own. Ethical AI use means facing the real, unresolved problems with how these tools get built including the fact that the people building them are now publicly asking each other to slow down and using AI in ways that don’t compound those problems, like prepping for paid expert work instead of trying to replace the experts entirely.
The Escalating Effort to Slow AI by Erle Norton, Wall Street Journal Article
Anthropic researcher resigns with warning about the dangers of AI development by Kaitlyn Huamani
