AI Foundations: Build a Bio That Teaches AI Systems Who You Are
AI Foundations: Build a Bio That Teaches AI systems Who you Are
Your bio is the first brick in a strong online identity, and it can have an outsized impact on how you show up in AI systems. Here’s how to build the one version of you that everyting else points back to.
Most professionals have a bio somewhere.
You may have two on LinkedIn – one long and one short. There’s probably an About page on your website. Or at least a paragraph that was supposed to become one. Maybe a speaker profile from something you submitted two years ago. A podcast guest intro you emailed a host at 11pm. A directory listing from a professional association you joined and forgot about.
None of them is wrong, exactly. But chances are good that none of them fully agree with each other either. This used to be okay. In fact, years ago, I would encourage folks to have multiple biographies – different versions of your heroic story told specifically for the intended audience.
But now…
Multiple different versions of you scattered around the web? That poses a bit of a structural problem. And you may not even realize you have that particular problem until you search for yourself – or worse, someone else searches for you – and find something unexpected looking back.
This edition is about fixing it. Specifically, it’s about building a single source of truth in the form of a canonical bio: one authoritative, deliberate, current statement of who you are, living on a page you own, that everything else in your online presence points back to and agrees with.
Identity infrastructure. And it’s the first brick in how people and machines come to understand who you are.
Why This Matters More Now Than It Used To
The way people find and vet professionals has changed.
Search engines have always pulled from multiple sources to build a picture of who you are. But AI-powered discovery systems – ChatGPT, Perplexity, Gemini, Copilot – have raised the stakes on consistency in a way that’s worth understanding.
AI doesn’t know you. When someone asks an AI assistant about you, or searches for someone who does what you do, the system isn’t pulling from a single authoritative file it keeps on you. It’s assembling you in real time from whatever signals it can find across the web – your website, your LinkedIn, your mentions in articles, your author profiles, your directory listings, your old bios that never got updated.
Then it looks for agreement.
When multiple sources say the same thing about you – same name, same role, same focus, same story – the system builds that picture with confidence. It can describe you clearly. It can recommend you.
When sources conflict, the system hedges. It qualifies. It says things like “details about her exact current roles and achievements can vary by context and source.” That’s not a neutral outcome. That’s the machine telling anyone asking: I’m not sure who this person is right now.
And here’s the part that a bit of a hiccup: you probably won’t know when this is happening. You’re not getting a notification that an AI described you wrong. You’re just out here – but not showing up when you should be.
A canonical bio, consistently signaled, is how you become the author of the loudest and clearest signal in that process. It’s how you give machines and people something authoritative to find.
What Canonical Actually Means
Canonical is a term borrowed from web development. Back in my SEO days, canonical was law. When multiple URLs contain the same or similar content, developers designate one as the canonical source – the original, the authoritative version, the one that counts. This became really important when the industry moved toward syndicating and repurposing content.
A canonical bio is the master record. The one you wrote deliberately, that reflects who you actually are right now, that lives somewhere permanent. Everything else – the LinkedIn summary, the speaker intro, the social profiles, the podcast guest bios – should all pull from your canonical bio, echo it, agree with it.
Not word-for-word, necessarily. But in substance: the same name, the same core description of what you do, the same framing of who you serve, the same story.
How does that look IRL? I’ll give you an example. In a few places online it says I’m an award-winning writer. My canonical bio lists the actual name of the award and the piece of work that won.
When your presence agrees with itself across platforms, that consistency is a signal in its own right. It says: this person has a clear, stable identity. Here is the source.
Why It Has to Live on Your Website
You might be thinking: my LinkedIn is more visible than my website. More people see it. Why not make that the canonical source?
Because LinkedIn is rented space.
Every social platform, every third-party profile, every directory listing exists on infrastructure someone else owns and controls. Algorithms change. Platforms get acquired. Features disappear. Terms of service get updated. What happens if LinkedIn gets enough pushback from users around AI scraping that ends up blocking search engines and AI systems from scraping the platform for information – and you’ve deemed it your single source of truth for all things? Frustration – that’s what happens. And probably an inconveniently-timed rebuild.
Your website is the only real estate you actually own. It’s the only place where you are permanently the authority – where nothing you say can be overridden by a platform decision you didn’t make.
When you establish your canonical bio on your own domain and signal it consistently across your ecosystem, you’re telling every search engine and AI system: this is the source. Start here. When in doubt, defer to this.
That matters when platforms change. It matters when a listing gets scraped wrong. It matters when an old piece of content resurfaces and conflicts with who you are now. Your website is the anchor. Everything else is a spoke pointing back to it.
What a Canonical bio Contains
Keep it straightforward. You’re answering five questions:
Who are you?
What do you do?
Who do you do it for?
What do they get from knowing you?
Where do they go next?
Write it the way you’d introduce yourself to someone worth impressing at dinner. Not keyword-stuffed. Not title-stacked. Plain language that a non-specialist can read and understand immediately, because AI systems, like humans, favor clarity over jargon.
Length doesn’t need to be precious. Two to four paragraphs is plenty. Long enough to be complete, short enough to actually get read.
Write it the way you’d introduce yourself to someone worth impressing at dinner. Not keyword-stuffed. Not title-stacked. Plain language that a non-specialist can read and understand immediately, because AI systems, like humans, favor clarity over jargon.
The Consistency Layer: Making It Work Across Platforms
Writing the bio is step one. Deploying it is step two. And this is where you want to be deliberate so you don’t leave value on the table.
Once you have your canonical bio, go through every platform where you have a meaningful presence and make sure the substance agrees.
LinkedIn is the most important external platform to align. Your headline, your About section, and your featured content should all reflect the same core identity as your canonical bio. They don’t have to be identical – LinkedIn has its own native voice – but they should tell the same story.
Your email signature, speaker profiles, and guest bios should pull directly from your canonical bio. When a podcast host asks for a bio, you send them yours. Not a custom-written version that introduces new language the machines have never seen before.
Your social profiles – wherever you’re active – should at minimum have a consistent name and a link back to your website. The link matters. It’s a signal. It tells systems: this profile and this website are the same entity.
Directory listings and author profiles are often overlooked and frequently outdated. Go find yours. Update them. Anywhere your name appears professionally online is a data point in how machines understand you. Outdated data points create friction.
The goal isn’t identical copy everywhere. It’s agreement in substance. Same name. Same core description. Same story. One master record that everything else reflects.
A Note on Keeping It Current
Your canonical bio is not a set-it-and-forget-it asset. It should evolve as your work evolves.
You don’t need to update it constantly. That would create its own kind of noise. But when something meaningfully changes – your focus shifts, you launch something significant, your audience changes, you move – update the canonical bio first, then let the other platforms follow.
Think of it as updating the master file before you update the copies.
The reason this matters: AI systems don’t have a fixed memory of you. They’re assembling you from current signals on a rolling basis – and the length of that “roll” depends on the platform itself. I call Perplexity the “fastest learner” because it pulls in real time and relies heavily on what it retrieves and not just what it already saw and was trained on. If your canonical bio still describes who you were two years ago, that’s what they’re working with. Keeping it current keeps the record accurate.
Tracking Professional Evolutions
When things happened matters as much as what happened. If your goal is to become recommendable in AI systems, keep in mind that machines don’t just want to know who you are. They want to understand the shape of your professional story – how you got here, what changed, why.
If your canonical bio says you’re an AI visibility engineer, but your old author bio is still floating around saying you wrote productivity books, the system has to reconcile those two things. Are they two different people? Did you pivot? When did that pivot happen?
Without a clear timeline, the machine fills in blanks. And it might fill them in wrong.
This is where the temporal layer comes in.
Your canonical bio should make clear: here is who I am now. And when you look at the consistency across your ecosystem – your LinkedIn, your website, your media kit, your speaker profiles – they should all reflect the same current version of you.
But here’s the important part: your professional history should be traceable, not erased.
If you wrote books five years ago, that’s real. It happened. But it should be contextualized and placed in your timeline so the system understands it as part
A Quick Story to Drive This Thing Home
At the beginning of February, I unpublished some old books I’d written. Not a big deal, right? They were from an earlier chapter – productivity guides, women’s empowerment content, work-at-home mom advice. Good work, but not where I am now. I pulled them off my Amazon shelf to update them, refresh them, get them right before re-releasing.
What I didn’t know was that Amazon was an authority node in my knowledge graph.
When I removed those books, I didn’t just clean up my shelf. I yanked a structural pillar out from under how AI systems understood who I was. And the systems did exactly what systems do when a signal goes dark: they went looking for a replacement.
They found one. Barnes & Noble. Everand. Old listings. Old descriptions. Old me.
On February 6th, Perplexity hedged.
Now, I’ve worked in the machine-visibility space long enough to know that if I change something, I need to see how it affected my story in machines. So, I did what I’ve done hundreds of times at this point – I searched “Who is Sorilbran Stone?” And I got this back:
“Because there isn’t a single canonical bio widely recognized across all sources, details about her exact current roles and achievements can vary by context and source.”
That’s the machine telling me, in plain language: I can’t find agreement. I don’t know who you are.
Three days later, it stopped hedging. It made a decision.
It decided I was someone I used to be. Productivity books. Small-business guidance for women. Atlanta-based author. All accurate – for 2016. Not for now.
In the screenshot above, the machine has decided it’s more confident in the older version of me – at least while it’s reweighting my identity without those decade-old Amazon nodes. Perplexity’s trying to decide if it’s still going to be the wind beneath my digital wings.
Ten years of professional evolution. Gone. Not because I did something wrong. Because I removed a source, and the AI filled the vacuum with whatever it could find.
This is called identity regression. And… it happens.
You better believe I got right on that canonical bio (you can check out mine here). It’s not that I didn’t have a biogrphy on my site – I did. But I didn’t have one that accounted for the stories that I may have been telling about myself in different places at different times. Or the story that Amazon had told about me – a story that was more important to the machine’s understanding of who I am than I realized.
Without the canonical bio tying everything together, the machine re-prioritized the signals that made up my identity and re-assembled my digital identity over the course of a few days to tell a completely different story about me. Wild, right? And fascinating. I truly am fascinated by its seemingly outsized response to that Amazon node disappearing.
But a truly canonical biography would have stabilized my identity, at least in how the machines saw me and weighted everything.
How to Know It’s Working
Search for yourself. Not just on Google – on ChatGPT, on Perplexity, on whatever AI tools your clients or peers are likely to use.
Ask: “Who is [your name]?” Ask: “Who does [what you do] in [your location or industry]?”
What comes back? Does it accurately reflect your current work? Does the story hold together? Or are you getting a version of yourself that’s hedged, outdated, or just slightly off?
If what you find doesn’t match what you’d want people to find, the canonical bio is where you start. Fix the source. Let the signal propagate. Then check again.
And this ☝🏽 is an example of me and the machine making peace. After I filled in the gaps with a canonical bio, of course. The experiment took 13 days altogether. I’m okay with this version of me being versioned and picked up by other AI systems. But if you’re in a competitive market and you and the machines are trying to figure out who you are for two weeks, that’s a heck of a lot of opportunity on the line.
Start Here
You don’t need a complete website overhaul. You don’t need a rebrand. You don’t need to become an SEO specialist or hire one.
You need one page, on your website, with one deliberate, current, accurate statement of who you are, written plainly, linked to consistently, updated when your work evolves.
Then you need everything else to agree with it. That’s the first brick. Everything else in how machines and people understand you builds on top of it.
If you don’t lay it deliberately, something else becomes the foundation by default. And you probably won’t know which version of you is holding up the wall.