Your Identity Is Your Infrastructure: Why AI Overlooks Senior Leaders
Senior leaders spend decades building stature, but AI does not automatically recognise it. In this episode of Hired to Lead, the podcast from Zeren, part of Renovata, host Kim Wardle speaks to Erin Lanuti, Co-Founder and CEO of Lilypath, about executive visibility in AI: why even well-known leaders are rarely recommended by AI tools, how thin profiles let others take ownership of their ideas, and what leaders can do to be represented accurately.
Key takeaways
- Stature is earned through roles, awards, books and interviews. None of it transfers to AI automatically.
- In Lilypath’s analysis, senior executives appeared in AI answers about 50% of the time, were described accurately about 75% of the time, and were recommended in only around 15% of answers.
- When a profile is thin, AI fills in the blanks, and people who understand AI signals can become the recognised authority on your topics.
- Misattribution is invisible to the person affected. There is no Google alert for how AI describes you.
- The fix starts with the foundation: a strong LinkedIn profile, then consistent content on credible third-party sites.
Who is Erin Lanuti?
Erin Lanuti is Co-Founder and CEO of Lilypath, which helps professionals understand and improve how AI platforms represent them. Her career began in PR at Burson and in entertainment marketing, including launch work on P&G’s Swiffer and Febreze. She later built Conversation to Commerce, an earned media measurement technology at Publicis, before leading innovation for Omnicom’s PR agencies, where she helped create Omni Earned ID and launched the AI Optics GEO offering in December 2023.
The launch of ChatGPT in November 2022 changed her focus. “I saw clear as day a warning,” she says. If AI gave a brand’s customers inaccurate information, “I’m completely unaware of it.”
What does Lilypath do?
Lilypath analyses a person’s LinkedIn profile section by section, scores how well LinkedIn’s AI understands it, and rewrites it so the person is read correctly. Its second product, Authority Graph, looks at what Claude, ChatGPT and Gemini say about a professional across five dimensions: search presence, what they stand for, understanding of their IP, misattribution and drift.
Erin tested it on herself first. Her original score was 51 out of 100. After making the changes and pushing her score from the low 80s into the 90s, she says her post reach rose 170%, and within weeks she was receiving speaking and interview requests. “The people that I want to see me are now seeing me because the AI is aligned with who I am.”
Why doesn’t executive stature carry over to AI?
Stature is a human signal. “It’s earned through shared experiences, it’s earned through awards, books, TV interviews,” Erin explains. “That doesn’t transfer to AI automatically, and I think we assumed it did.”
Lilypath’s analysis of senior executives found they appeared in AI answers about half the time. When they did appear, accuracy was around 75%. But on average, AI actually recommended them in only about 15% of answers.
What happens when AI fills in the blanks?
Many senior leaders assume the people who matter already know them. Erin argues that this is the risk. “If you have a thin profile, the AI fills in the blanks.”
Others who understand AI signals are moving into that gap. “They don’t have the stature, but they’re getting the AI recommendations because they’ve filled in all the blanks.”
It happened to Erin. Her weekly report showed Claude and Perplexity assigning IP she had trademarked to someone else. “The scarier thing about it is you’re invisible to this. It’s not like you can set a Google alert as a professional.”
How badly can AI get an executive wrong?
In one case, Lilypath ran a senior executive through its system. His visibility was reasonable, but accuracy was badly off. AI was confusing him with someone of the same name who was under federal indictment.
The usual response is that people who know the executive will know better. Erin’s view is that this is changing. “As agentic moves faster and an agent increasingly talks to an agent to collect the information, there is no human in the loop.” In that setting, she says, the perception is the reality.
What is the real cost of being invisible to AI?
The cost is opportunity you never see. “You’re never gonna know the job that you were filtered out of. You’re never gonna know the speaking opportunity.” Erin sees the same pattern at every level, from graduates applying to hundreds of roles to experienced professionals looking for their next position, because the way people find and assess candidates has changed.
How can leaders improve their executive visibility in AI?
Erin’s advice is to fix the foundation before posting more.
- Start with LinkedIn. Erin describes LinkedIn as now the leading source for professional citations in AI answers. Make sure your profile states clearly what you want to be known for.
- Align what you post with your profile. When posts match the profile, LinkedIn’s systems treat you as an expert on the topic and give you more reach.
- Find the gaps. Check which topics AI associates with you, where you are weak, and who is being credited instead.
- Publish on credible third-party sites. Blogs, Substack, podcasts and your own website all add signals. “You want consistent, deep signals across platforms to make sure that the AI understands who you are.”
- Check the citations. AI is confident even when it is wrong, and it still makes up sources.
Why does identity matter in an AI world?
“We believe that your identity is your infrastructure,” Erin says, “and that you should have control over that in the AI world.” She is concerned about people being misrepresented and fake candidates appearing in interviews, and wants better tools that let people see and influence how they are portrayed.
For boards and investors, the same questions now apply to leadership teams. How AI describes a CEO or a newly appointed executive is part of how customers, investors and future hires first meet them. We explore how AI is reshaping leadership roles in Dissolving disciplines: who owns Product, Engineering and Design now that AI blurs the lines and Turning AI into EBITDA.
Leading at speed: the speedboat method
Moving from Omnicom to a founder role has meant wearing many hats and making decisions without waiting for approvals. At Omnicom, Erin used what she calls speedboats: inviting one agency to test a new idea fast, then share what it learned so the rest could scale it. Lilypath runs the same way, placing many small bets and backing the ones that gain traction.
Quick-fire questions
- Visibility or accuracy? Accuracy
- Prompt or proof? Proof
- One answer or a pattern over time? Pattern over time
- Human stature or machine authority? Both
- Post more or fix the foundation? Fix the foundation
- AI optimist or AI realist? Optimist
Frequently asked questions
Does a strong reputation guarantee AI will recommend you?
No. In Lilypath’s analysis, senior executives were recommended in only around 15% of relevant AI answers, even though they appeared in about half of them.
Does LinkedIn affect what ChatGPT and other AI tools say about you?
According to Erin Lanuti, yes. She describes LinkedIn as the leading source for professional citations in tools such as ChatGPT, Claude, Perplexity and Gemini.
What is AI misattribution?
Misattribution is when AI credits your work, ideas or IP to someone else, or confuses you with another person. It usually happens without you knowing.
What should a senior leader do first to improve executive visibility in AI?
Fix the foundation: a clear LinkedIn profile that states what you want to be known for, then consistent content on credible third-party sites.
Watch the full episode
Watch the full conversation with Erin Lanuti on YouTube. To find out more about Lilypath, visit lilypath.com.
Hired to Lead is the podcast from Zeren, part of Renovata, hosted by Kim Wardle. Each episode goes beyond the CV to explore the journeys, pivots and defining decisions of leaders at the top.
Full transcript
Read the full transcript of Hired to Lead with Erin Lanuti
Kim Wardle: Hi, and welcome to another episode of Hired to Lead, a podcast brought to you by Renovata. I’m Kim Wardle, and we don’t just place world-class leaders, we also give you insight into their journeys, the moments that tested them, the pivots that they didn’t see coming, and the decisions that defined everything that followed: the stories that don’t make it onto a CV, a portfolio or a LinkedIn profile. We believe our community deserves direct access to the people doing it at the highest level. Honest, unfiltered conversations on the doubts, the defining moments, and what it really takes to lead at the top. This is Hired to Lead, and I’m excited to jump into it with today’s guest, Erin Lanuti, CEO and founder of Lilypath, who, along with her two co-founders, is striving to build something the market has never seen before. Hi, Erin.
Erin Lanuti: Hi, how are you?
Kim Wardle: I’m great, thanks. Awesome to have you on here.
Erin Lanuti: Thank you for having me. I’m excited to have this conversation with you.
Kim Wardle: So we first met just over a year ago when you attended my innovation dinner in New York. At the time, you were leading innovation at Omnicom, and I can’t believe how much has changed in such a short amount of time, which I’m really excited to share with our listeners. I’d love for you to start with your journey and how you got to where you are today.
Erin Lanuti: It’s been a journey, let me tell you. In my past, I have walked penguins down Fifth Avenue. I have had Muppets light the Christmas tree in Washington, DC, and I have launched the first avatar spokesperson for an Olympic Games. So that’s where I started my career. Imagine going from that to CEO of Lilypath today. That’s quite a gap, so I hope you have a cup of tea while I walk you through how I got all those things together.
I started my career at Burson, learning all the ropes: consumer launches, crisis, B2B, event marketing, and that was fabulous. Then I switched to another firm where I worked on P&G’s new-to-the-world products, so Swiffer and Febreze. In launching those, we had huge budgets, hundreds of millions of dollars. But at the time, we were struggling to make an emotional connection with consumers, which brought us into entertainment marketing. That was something I worked with the agency and P&G to spearhead, ultimately putting, believe it or not, Jessica Simpson on the cover of Rolling Stone holding a Swiffer WetJet. That was a defining moment.
From there, I moved into entertainment marketing and started my own agency. We represented all the studios, basically aligning studios and companies and doing all those deals. I did that for several years, and then was lured back to MSL, which is part of Publicis, to work on their largest account at the time, which was UL, Underwriters Laboratories. Most people have no idea, but it’s on almost every device you have: a little circle, UL, and it’s a safety seal.
For that, I ran everything: paid, owned, earned and creative. It was a fantastic way to understand all the different disciplines, figure out how I could interconnect them and better understand measurement, which is how I built a technology for Publicis called Conversation to Commerce. That technology was pivotal because earned media didn’t really have a way to be measured at the time. Clients would always ask, “What’s the ROI on that? Is it impressions? Is it this? Is it that?” So I crafted this technology that allowed us to take earned media and convert it into the equivalent of paid ads, so we could turn that content into an ad and target it to whoever we wanted. It was a huge deal at the time. We scaled it to, I think, over 50 countries, and that became the core of how they started to measure earned impact.
Fast-forward from there, I was asked by Omnicom to lead innovation for their PR holdings, and Omnicom has the most PR agencies of any company in the world. Being able to take what I had learned at Publicis and take it to the next level with Omnicom’s data and tech was a dream. When I went there, we looked at all the data they had and what they were doing in paid media and said, “We could reapply a lot of the paid data to earned data to track brand lift and sales lift.” That became the impetus for Omni Earned ID, which is now the gold standard of measurement across the agencies.
So that’s a very long, linear journey, and things were going great until November 2022, when ChatGPT launched. Everything upended. Everyone had to figure out what it is, how to train their people on it, how to apply it and what the use cases are. In doing all of that, one thing was really clear to me: AI search and its implications were going to be huge.
In September 2023, I started figuring out what this was going to mean, because I saw clear as day a warning. If you’re a brand and someone asked ChatGPT about your product or service, and it came back with something completely inaccurate that could lead to misuse of a product and harm somebody, you’re completely unaware of it. I thought, “This is a huge brand risk.” So I started writing memo after memo and looking for technologies that could help us understand what AI was saying at scale so we could protect our clients. Those were the early days of what we now call GEO or AEO. In December 2023, we launched our GEO offering, AI Optics, with a company called Bluefish, which was the first technology at the time that could handle enterprise scale.
Everything was going well, but as things were changing and Omnicom was going through the transition everyone knows about today, I was thinking, “Maybe now is a good time to go out on my own. Maybe everything I’ve learned in GEO and AI transformation would be better suited on the other side.” Then I was approached by Lilypath and the technology they had created. They said, “We have this LinkedIn thing. We want you to look at it.” I thought, “LinkedIn is a walled garden. I know everything I need to know about LinkedIn.” They said, “Just look at it.” So I said, “Fine. Run it on me and let’s see what we find.”
I was shocked. In the earliest days of Lilypath, it was pretty harsh. They gave you letter grades, and mine was an F. I got 51 out of 100. I said, “How can this be? Is my LinkedIn profile really this bad? I do this stuff every day.” I thought, for sure the tech must be wrong. But I started digging into it and realised LinkedIn had put all these AI algorithms on their platform without telling us. Everything you were doing on the platform was being gated by AI, and it was exactly the same parallel I saw on the AI search side. Now I was seeing it on LinkedIn.
Kim Wardle: So essentially, Lilypath’s product scrapes your LinkedIn profile to give you a score?
Erin Lanuti: That’s right. We analyse your LinkedIn profile and look at every single aspect of it. We break it into pieces, your headline, your about section and so on, and analyse it through the lens of LinkedIn’s AI, which is reading it, to make sure it’s reading it correctly. Are you accurately portrayed? Does the AI understand who you are? Every piece is broken apart and scored, and then the part people love most is that we rewrite it so you have a higher score and can be seen.
In my case, starting at 51, I ran it on myself and ended up in the low 80s. Because I’m a Type A overachiever, I thought, “How do I get to the 90s?” So I made the changes again and reran it. What I saw was that my reach went up 170% when I posted. Then I started seeing more follower views. But these are all metrics, and I wanted to know what that tangibly meant. Within a couple of weeks of making the changes, I started getting requests: “Will you speak at this conference?” “Do you want to do this interview?” I thought, “The people I want to see me are now seeing me because the AI is aligned with who I am.” That was when the penny dropped for me on the real importance of this.
Kim Wardle: Incredible. Let’s slightly change course, because you once said to me that people spend their careers building stature, but that stature doesn’t automatically transfer to machines. Unpack that for us.
Erin Lanuti: This has been one of the biggest aha moments for me. If you think of the top CEOs in finance, arguably they have the highest stature. But stature is a very human thing. It’s earned through shared experiences, awards, books, TV interviews and roles. That doesn’t transfer to AI automatically, and I think we assumed it did.
We have another product called Authority Graph, which analyses people against what Claude, ChatGPT and Gemini say about them. We look at it through five dimensions: search presence, whether it knows what you stand for, whether it understands your IP, whether there are misattributions of your IP, and where it is drifting. So we’ve built technology that is all about professional reputation.
Looking at these executives, what we found, and this is going to blow your mind, is that on average they show up about 50% of the time. These are the biggest names out there, and they appear only half the time when you search for who they are and the topics they stand for. When you look deeper, accuracy when they do show up is about 75% plus, which is pretty good. But the biggest issue is that only 15% of the time, on average, does the AI actually recommend them in its answers.
We hear all the time from top CEOs, or people who think they have all this stature, “I don’t need to bother with that. The people who know me, know me, and they’ll know who to call.” But the problem with AI is that if you have a thin profile, the AI fills in the blanks. It’s a risk to have a thin profile. And the bigger risk we’re seeing, which is so crafty, is that people who understand AI, signals and building authority in AI are now capitalising on the fact that these people have thin profiles, and they’re becoming the authority on the topics.
They don’t have the stature, but they’re getting the AI recommendations because they’ve filled in all the blanks, and in AI’s eyes they own that. And you’re completely invisible to it. It’s the modern-day equivalent of Who Moved My Cheese: someone is taking your IP and running with it.
That was something we’d been digging into with these professionals. Then, in my own report, which we’ve been running weekly for months, I logged in one week and saw a huge red line. I thought, “What is that?” It was Claude and Perplexity taking IP that I had trademarked and owned and assigning it to somebody else. Now this was really real; it was impacting me too. But the scarier thing is that you’re invisible to this. It’s not like you can set a Google alert as a professional. That’s why we believe so strongly in helping professionals understand what’s being said and helping them be visible and accurately portrayed in AI.
Kim Wardle: That’s fascinating and also scary. Can you give us a real example of where AI is getting a prominent executive seriously wrong?
Erin Lanuti: Unfortunately, this one was also pretty scary. We ran a top executive through the system. His visibility was around 68%, which is okay, but his accuracy was way off. When we dug in, the AI was misattributing his name. There was another person with the same name, except that person was under federal indictment and accused of mafia ties.
Kim Wardle: Oh my goodness.
Erin Lanuti: When we looked at it, it was about as bad as it could possibly be. If I showed you the dashboard, everything was red. But again, this is something people aren’t aware of. You don’t know how AI is portraying you. And again, what we hear is, “That’s bad, but I’m so prominent that the people who know me will know that’s not really me.” But as agentic AI moves faster and an agent increasingly talks to an agent to collect the information, there is no human in the loop. The perception is the reality in that instance, so this becomes even more important to fix.
Kim Wardle: That’s a very different kind of risk from just not showing up at all. So when did this stop being someone else’s problem and start being personal to you?
Erin Lanuti: There are so many dimensions to this problem. If I look at early-career people, college students applying to hundreds of jobs and not getting them, that’s a structural problem. They’re not getting through the AI systems and being seen. If I look at friends and colleagues at firms I’ve worked with over the years who’ve lost jobs, they’re all facing the same thing. They’re not being found or seen because the playbook has changed. The way people find jobs has changed.
And if I go to the other extreme, the executives I’ve been talking about, this is invisible opportunity lost. You’re never going to know the job you were filtered out of. You’re never going to know about the speaking opportunity or the media interview. It just won’t appear. That’s the biggest shame of it all, because this is already happening, and people aren’t aware of what’s at stake and what’s being lost every day.
Kim Wardle: So can you walk me through what someone can do to fix it?
Erin Lanuti: Of course. A lot has changed. Let’s start with LinkedIn. LinkedIn is now the number one source for professional citations. Said another way, what happens on LinkedIn doesn’t stay there anymore. It’s fuelling the responses of ChatGPT, Claude, Perplexity, Gemini and others.
To fix these things, the first thing you need to do is fill in the gaps for AI. I’ll give you a case study: the example I gave you where someone is taking my IP. We ran the technology on it and asked what messaging they were trying to have attributed to them. Then we built a consistent initiative to make sure I was getting credit for those things.
It starts with LinkedIn: making sure you have a very strong profile and that what you want to be known for comes through in it. Then, when what you post aligns with what your profile says, you get more reach, because LinkedIn’s AI systems say, “Erin is an expert in authority intelligence, so when she posts about it, I’m going to give her more reach.”
Then, when you go to Claude, ChatGPT and the others, our authority intelligence shows you specifically who’s saying what, where you’re weak and which topics are your weak spots. If I start writing blogs or Substacks or doing podcasts that say, “I’m the inventor of authority intelligence, and here’s what it means and why it matters,” and that’s published on credible third-party sites, it boosts my AI authority. When I post on my own brand site or Substack, all of those things help make sure your signals align. You want consistent, deep signals across platforms so that AI understands who you are and why you’re the best answer to a question.
Kim Wardle: And what do those cases tell us about trust when AI gives a specific, confident but inaccurate answer?
Erin Lanuti: The amazing thing to me with AI, and I face it all the time with Claude, ChatGPT and all of them, is that it is so confident in the answers it gives you. It still makes up citations to this day. So checking the citations, whether they exist and whether they’re real, is increasingly important.
Trust is one of the things that really concerns me in AI right now. Lilypath’s mission is to make sure people are visible and accurately portrayed in an AI world. We believe that your identity is your infrastructure, and that you should have control over that in the AI world. I’m extremely concerned by some of what we’re seeing: people being misportrayed, fake people showing up for interviews. It’s real, and it’s only going to continue to scale. So I think there needs to be more visibility into how you’re portrayed, and more tools and technology that let you influence how you’re portrayed online, so that you have control over your identity.
Kim Wardle: And hopefully Lilypath can give us that.
Erin Lanuti: I certainly hope so. We’re trying really hard.
Kim Wardle: I’m going to change course slightly, because I’m really interested in you as a leader. You’ve moved from one of the biggest agencies in the world to building your own company. How has that changed you?
Erin Lanuti: It’s changed me a lot. There are very few people to pick things up, so you have many roles on any given day. Ironically, all the pieces of my career now seem to come together and play a more meaningful role. Who knew I’d have a website development skill now, alongside PR and everything else? It’s been extremely fun and exciting. At an agency you work crazy hours and pour your heart and soul into everything. The cool thing now is that you’re doing it for yourself. And the speed is so fast; you don’t have to wait for decisions and approvals to come down. You can just say, “Go.” That’s been interesting for some of the people we work with. They say, “Really? That’s it?” And I say, “Yes, go. Just do it.”
Kim Wardle: When we first met, you told me about your speedboat methodology for innovation. I tell a lot of people about it. Could you tell our listeners a little more about it?
Erin Lanuti: Of course. I have always believed that in large organisations you can get bogged down. You’re waiting for approvals and you can’t move quickly. So I developed a methodology I called speedboats, which is a way for people to move quickly in a big organisation. At Omnicom, I would say, “Here’s a big change happening in the industry. Which one of you wants to raise your hand and be the speedboat to go out and test it?” I would make lots of little bets. The good thing for whichever agency raised its hand was that it would be first, so it had the advantage. But in a holding company, the key part was that they had to come back with the learnings, so we could all understand what was happening and be better prepared to scale it across every agency.
The same thing holds in a startup. You place lots of little bets, look at the signals to see what’s gaining traction, and then go deeper once you learn something. So the speedboat methodology is alive and well and thriving at Lilypath.
Kim Wardle: I love that, and I’m excited to follow you on this journey and see what’s next. For now, I’m going to throw some quick-fire questions at you. First instinct, no overthinking. Ready?
Erin Lanuti: I’m ready.
Kim Wardle: Visibility or accuracy?
Erin Lanuti: Accuracy.
Kim Wardle: Prompt or proof?
Erin Lanuti: Proof.
Kim Wardle: One answer or a pattern over time?
Erin Lanuti: Pattern over time.
Kim Wardle: Human stature or machine authority?
Erin Lanuti: Both.
Kim Wardle: On the fence on that one. I’m not sure that’s allowed, but I’ll give it to you. Post more or fix the foundation?
Erin Lanuti: Fix the foundation.
Kim Wardle: I definitely need to take that one on board. AI optimist or AI realist?
Erin Lanuti: Optimist.
Kim Wardle: Erin, this has been brilliant. Unfortunately, we’ve run out of time. I know Lilypath is doing something really exciting in this space, and I want our listeners to know more. Listeners, please check out Lilypath’s brand-new website at lilypath.com. And if you want to hear more from Erin, keep an eye on her LinkedIn profile, as she’ll be doing more talks on AI and Lilypath’s product features. I hope you’ve enjoyed today’s episode. I know I have. Please subscribe, leave us a review and share it with someone in your network who you think would get something out of it. We’d love to keep this community growing. This has been Hired to Lead, brought to you by Renovata. I’m Kim Wardle. Erin, thank you so much.
Erin Lanuti: Thank you so much.
This article was first published on renovata.com.
Hired to Lead is a podcast by Zeren, part of Renovata. We partner with ambitious businesses on leadership search and the hires that drive growth. Get in touch to discuss your next leadership hire.
