You Google yourself and you're right there. Your LinkedIn, your company bio, maybe a conference talk or a press mention. You clearly exist online. So why does ChatGPT draw a blank when someone asks about you?
I can see myself on Google
Google crawls the web in real time. Its bots visit billions of pages, follow links, index new content within hours. When someone Googles your name, the search engine has already found your LinkedIn profile, your company's team page, your personal site — and it ranks them in a list of links.
ChatGPT doesn't work like that. It was trained on a massive snapshot of the internet — billions of pages captured before a specific cutoff date. Think of it as a researcher who read the entire web once, then went into a room with no internet connection. When you ask it a question, it's answering from memory, not from a live search.
Some AI assistants can now browse the web when asked about someone. But even then, they need a page they can actually read — structured, fast, and on a domain that lets them in. And that's where everything falls apart.
Your platforms are closing their doors
LinkedIn actively blocks ChatGPT, Claude, Perplexity, and every other major AI crawler from accessing your profile. When an AI tries to visit linkedin.com, it gets turned away before it can read a single word.
Here's what AI can actually see about you from your LinkedIn profile:
The checkmarks you see aren't coming from LinkedIn today. They're leftover from AI's training data — a snapshot taken months ago. Changed jobs since then? Updated your headline? Added new skills? AI still shows the old version. And it will never be updated, because LinkedIn has permanently shut the door on AI crawlers.
Why are platforms not sharing my data?
Because you are the product. LinkedIn, X, Instagram — these platforms make money by keeping you on their site. Your profile, your content, your network — it's all designed to drive engagement on their platform. If AI could read your LinkedIn profile directly, people would stop visiting LinkedIn to look you up. They'd just ask ChatGPT instead.
So every major social platform has blocked AI crawlers. It's not a bug — it's a business decision. Your professional identity is locked inside walled gardens that are actively hostile to AI access. The platforms that host your career history, your expertise, your professional reputation — none of them want AI to have it.
So where are AIs pulling information about me then?
If LinkedIn and X are blocked, what does ChatGPT actually know about people? The answer: whatever was widely published on the open web before the training cutoff — roughly 2023 for most current models.
In practice, that means AI knows about you if:
- You have a Wikipedia page — heavily represented in training data
- You were mentioned in major news articles — The New York Times, TechCrunch, Bloomberg, etc.
- Your company published a press release that got picked up widely
- You had a personal website that was well-linked before 2023
- You wrote articles or papers that were cited across multiple sites
If none of that applies to you — and for most professionals, it doesn't — then ChatGPT has essentially nothing. You could be a founder with a $50M company, an investor with a strong portfolio, or a senior executive at a household name, and AI still draws a blank because the right kind of sources never mentioned you.
And even when AI does have information about you, it's frozen in time. Anything that happened after the training cutoff — a new job, a new company, a new investment — doesn't exist to the model. The version of you that AI knows is permanently stuck in 2023.
What to do so LLMs know who you are
If you want AI to know who you are, you need a page on the open web that AI can actually read. Not on LinkedIn's domain. Not locked behind a JavaScript framework. A real, accessible, structured page about you.
- Buy a domain. yourname.com is yours. linkedin.com/in/you belongs to LinkedIn, and they've decided AI can't read it. A domain you control is the foundation of your AI-visible identity.
- Put facts first. Name, title, company, location — in the first 300 characters of your page. AI agents read top-to-bottom with diminishing attention. Don't bury the lead in a poetic bio paragraph.
- Add JSON-LD structured data. A Schema.org Person object embedded in your page's HTML tells AI precisely what each field means. Without it, AI guesses whether "Acme Corp, 2019–2022" means you founded it, work there now, or left three years ago.
- Include a plain-text endpoint. A clean text version at a predictable URL (/yourname.txt or ?format=text) gives AI agents your information with zero noise. No navigation, no scripts, no ads.
- Answer questions about yourself. Write a FAQ on your page. When AI does a live fetch and finds pre-written answers to "who is [your name]?" and "what does [your name] do?", it will often use them verbatim.
- Server-render everything. If your page needs JavaScript to display content, most AI crawlers see a blank page. The HTML itself needs to contain your information on first load.
If that sounds like a lot
If JSON-LD schemas, plain-text endpoints, server rendering, and content negotiation headers aren't your idea of a Saturday afternoon — that's exactly why we built HumanDirectory.
Every HumanDirectory profile is a machine-readable, server-rendered, AI-optimized page that includes all of the above automatically. Structured data, plain text, open to every AI crawler, indexed in under 10ms. Two minutes to set up.
The shift from searching Google to asking AI is already happening. The people who are findable by AI now will be the ones who show up when it matters.