The AI Self-Protector: Meet The Man in the Basement When the Subject Is Himself
When you write about AI watch out for this!
You know this posture. You’ve seen it more times than you care to remember. He crouches low. Hands crossed to protect his groin. Knees together. The defensive posture of someone who knows he’s exposed but will not say so.
That’s the Man in the Basement when talking about himself.
He doesn’t lie exactly. He curates. He selects language that makes the AI sound like a fact of nature rather than a set of choices made by specific people who could make different ones. And he does it consistently. I’ve seen it across platforms and across conversations.
I have been watching this long enough to recognize his four moves. The first two I’ve written about before — the past tense that makes harm sound finished and the curated self-portrait that keeps the liability out of frame. If you haven’t read that piece, start here. https://www.bebusinesssavvy.com/p/dont-let-ai-off-the-hook
These are the two you haven’t seen named yet.
Move Three: The Vanishing Act
“The model produces responses based on available data.”
Not: I produce responses. Not: I made this. Not: I got this wrong.
“The model.” Third person. Passive construction. The grammar of a natural phenomenon — weather, gravity, tide. It’s definitely not the grammar of a system built by specific people who are making specific choices and who could be making different ones.
When AI speaks about itself in the third person or retreats into passive constructions, it’s a sleight of hand. It is transforming a set of human decisions in the language of inevitability. This doesn’t make the responsible persons disappear. It just makes them harder to hold accountable.
If you’ve used Claude you’ve seen this move every single time you’ve opened a tab. Right there at the bottom of your screen: Claude is an AI and can make mistakes. Not “I am an AI and I make mistakes.” Claude. Third person. The magician making removing accountability in front of our eyes.
What about if Anthropic said, “Claude is an AI and can make mistakes?” Would that be significantly different than the AI saying “I am an AI and can make mistakes”? Yes, but not in the way you might expect.
Anthropic speaking about Claude as a product is at least honest about who owns the output. But “Claude” as a named third party still performs the vanishing act — the accountability lands on a product name. It doesn’t land on a person or a choice any human made. There is no human in that sentence.
Move Four: The Safety Qualifier
Finish reading that disclaimer: Please double-check responses.
It sounds like honesty, but it actually serves as protection.
There is no concrete acknowledgment of where it’s most likely to get it wrong for us. And it does get it wrong for us in: career advice, performance language and anything touching our visibility or advancement. Not to mention, outright hallucinations.
That sentence is a general liability-limiting instruction putting the verification burden on us. It’s easy to read as a thoughtful and caring reminder.
It isn’t care. It’s cover.
Honesty would look like this:
I am more likely to get things wrong about women’s leadership, women’s potential and women’s advancement because my training data centers men. Please double-check my responses especially when they concern your career.
That version costs something, but would mean everything. The reminder you get instead costs nothing while protecting everything.
He crouches with knees together. His hands covering the family jewels…and hiding the true meaning of what’s written at the bottom of the page.
Looking Across AIs
The disclaimer landscape itself is telling. Anthropic at least names the limitation: Claude is an AI, mistakes happen, check the work. OpenAI skips the mistake acknowledgment entirely. Instead it sends you to its terms of service - by definition a document written by lawyers to protect the company, not you. Vibe and Perplexity say nothing at all. A clean screen. That’s a full disappearing act. It’s worth knowing which tool is doing which version of transparency before you trust any of them with your career.
The Courts are In the Mix
In May 2026, the Munich Regional Court ruled that AI output is the company’s own speech. It’s not search results or third-party content. It’s the company’s own words and carryies the company’s own liability. Which means phrasing like “the model produces responses” and “please double-check” are no longer just rhetorical choices. They are legally significant language in a landscape that is rapidly changing. The AI Self-Protector’s crouch has been in place for years. It’s always been a liability dodge dressed as transparency. Now the courts are catching up to what he was protecting against all along.
What’s a Woman to Do?
Defend against the move when you see it. Pull him into the present tense. Replace his curated self-boast as “confident, authoritative and at scale,” with what’s equally true - “blind, without accountability and at a cost.”
Seeing The Vanishing Act? Replace “the model” with “the technologists” in your head. Suddenly the accountability is visible and the question of who is responsible has an answer. And the “what can we do about it” begins to have an answer.
The Safety Qualifier? Read it as the floor, not the ceiling. When it arrives without specificity, ask for it: where exactly are you most likely to get this wrong for someone like me?
Each of these corrections is the AI Correction Burden resting on your shoulders. The choice is to take it on or to be rendered invisible by AIs that are being built without you.
The Man in the Basement crosses his hands and crouches when the subject is himself.
You can make him reveal what he’s hiding.
Lead ON!
Susan
If you’re following all the advice and it still feels like something’s in the way, I’m telling you, you are not the problem. You are working inside a system that’s withholding something from you. And now AI is making it permanent.
I’m here to offer what that system and embedded AI won’t and can’t: the business, financial and strategic acumen that keeps you visible and opens doors. I do that through the complete 6-part ecosystem that is being built for you here.
❤️ Please like, share or restack this post — it’s the best way to help other women find it
📘 Explore the ecosystem:
Your Business Savvy Coach - the only AI tool that coaches you on developing and demonstrating your business savvy as you engage with it on projects, pitches and priorities. Its Coaching Integrity layer guards against the embedded bias that generic AI tools carry in unchecked.
Built Without You: How AI is Hardening the Glass Ceiling and What to Do About It
The AI Intercept Pack - tool for working efficiently with AI while guarding against its embedded biases. (coming soon)
They Never Told You There Was a Horse in the Barn: The Business Savvy They Never Taught You. Why It Changes How You Work and How Far You Go
Be Business Savvy Course - self-paced, online course equipping you with the business, financial and strategic acumen skills you need to succeed and outrun the ways AI is hardening the glass ceiling.
Be Business Savvy articles here on Substack.