Last month, an event planner called us in a mild panic. She'd just watched the keynote at her industry's biggest annual conference and realized, about fifteen minutes in, that the "AI expert" on stage was essentially reading ChatGPT's greatest hits back to an audience of people who already used ChatGPT daily. The speaker had looked great on paper. Impressive bio. Professional headshot. A book with "AI" in the title. But when it came time to deliver actual insight, there was nothing there.
This happens constantly. The AI speaking circuit has become saturated with people whose primary credential is that they talk about AI, not that they've actually built anything, deployed anything, or helped real organizations navigate real decisions. The gap between "sounds credible" and "is credible" has never been wider.
Here's how to tell the difference before you sign a contract.
The Credentials That Actually Matter
Let's start with what doesn't automatically qualify someone to speak about AI at your event: a LinkedIn bio that says "AI Thought Leader," a podcast about technology trends, a consulting practice with "AI" in the name, or a book deal. These things are easy to acquire. They tell you someone is good at marketing themselves, which is a real skill but not the one you're paying for.
What you actually want is evidence that this person has been in the room when hard decisions were made. That they've had to explain to a CEO why a project failed. That they've built something that shipped. That they've been wrong in public and learned from it.
The most credible AI speakers typically fall into a few categories:
Builders have created AI systems that real people use. This might mean founding a company, leading an engineering team, or architecting systems at scale. The key word is "real." A demo that impressed investors is not the same as a product that survived contact with actual users.
Deployers have implemented AI within existing organizations. They've dealt with the messy reality of integrating new technology into legacy systems, resistant teams, and uncertain ROI calculations. These speakers often have the most practical insights because they've lived the problems your audience is facing.
Researchers work at the frontier, publishing peer-reviewed work or leading teams at institutions that shape the field. True researchers are often more measured in their claims and more comfortable saying "we don't know yet" because they understand how much remains uncertain.
Experienced observers have spent years covering the field, interviewing practitioners, and synthesizing patterns. Good journalists and analysts can offer valuable perspective, but they should be transparent about the nature of their expertise.
When you're evaluating potential speakers, the first question to ask yourself is which type of expertise your audience actually needs.
How to Verify What's Real
Bios are marketing documents. Everyone's bio makes them sound impressive. Your job is to look past the polish and verify what's actually there.
Check their work history, not just their titles. A "former Google AI researcher" might have been a contractor for eight months or might have led a core team for a decade. Both could technically claim that description. Look for specifics: team size, project names, outcomes that can be independently verified.
Search for their name beyond their own website. What comes up? Are they cited in news stories by journalists who actually cover AI? Do working researchers reference their work? Have they testified to Congress or advised actual policymakers? These are things that are hard to fake because they require external validation.
Look at their publishing history. Not blog posts on their own site, but peer-reviewed papers, contributions to real publications, or technical documentation. If someone claims deep technical expertise but has never published anything technical, that's worth noting.
Watch their previous talks. This is crucial and surprisingly few event planners do it thoroughly. Find three or four videos from different events over the past two years. Are they saying the same things with the same slides? Are they making predictions that have already proven wrong? Do they handle Q&A with ease or do they deflect and generalize?
Talk to people who've worked with them. Not just event planners who've booked them, but people who've worked alongside them professionally. A speaker might be great on stage and know very little about the subject. Colleagues will tell you what someone actually knows.
Red Flags That Should Make You Pause
Some warning signs are subtle. Others are screaming at you in neon. Here are the ones we've learned to spot:
Credential inflation. Phrases like "one of the first" or "among the leading" or "pioneering" are often doing heavy lifting to obscure a thin resume. Specifics inspire confidence. Vague superlatives should make you ask follow-up questions.
Expertise that appeared overnight. If someone was speaking about blockchain in 2021, leadership in 2022, and AI in 2023, they might be a talented generalist or they might be a trend-chaser. Look for sustained engagement with AI specifically, ideally stretching back before the current hype cycle began.
All prediction, no track record. Anyone can speculate about the future. The question is whether their past predictions panned out. A speaker who made bold claims in 2022 about how AI would transform business by 2024 should be able to point to where they were right and, just as importantly, where they were wrong.
Inability to explain the basics. True experts can make complex ideas accessible without oversimplifying. Ask your potential speaker to explain how large language models work in terms a curious teenager could understand. If they can't do it, or if they immediately pivot to buzzwords and brand names, they may be working from a shallow understanding.
Resistance to questions about their work. When you ask for specifics about projects they've led or systems they've built, a credible expert will give you more than you asked for. They'll light up because they love this stuff. Evasiveness is a bad sign.
We've written at length about warning signs to watch for in AI speakers, but the core principle is simple: real expertise loves scrutiny. Fake expertise avoids it.
Questions to Ask During the Vetting Call
When you get a potential speaker on the phone, treat it like a job interview because that's essentially what it is. Here are specific questions that separate genuine experts from skilled performers:
"What's something you believed about AI two years ago that you've changed your mind about?" Real practitioners are constantly updating their mental models. Someone who has all the same opinions they had in 2023 isn't paying attention.
"What's the most common misconception about AI that you hear from audiences like ours?" This question reveals whether they've actually engaged with audiences in your industry and understand their specific knowledge gaps and concerns.
"Can you walk me through a project where AI didn't work as expected? What went wrong?" Failure stories are gold. They demonstrate humility, real experience, and the kind of practical wisdom that comes from actually doing the work.
"What would you tell our audience that they probably don't want to hear?" Great speakers are willing to challenge their audiences, not just validate them. If someone will only deliver good news, they're probably more interested in booking fees than in being useful.
"How do you stay current?" The field moves fast. Someone who can't describe their learning practice, whether that's reading papers, building prototypes, or participating in technical communities, may be coasting on outdated knowledge.
Industry-Specific Expertise vs. General AI Knowledge
There's another dimension to credential evaluation that event planners often overlook: the difference between knowing AI and knowing your industry.
A brilliant machine learning researcher might have deep technical knowledge but no idea how healthcare compliance works, or why manufacturing supply chains are complex, or what keeps financial services executives up at night. Conversely, someone with twenty years in your industry who's spent the last three years deeply focused on AI applications might offer more relevant insight than a famous technologist.
The best speakers for your specific event usually have some combination: enough technical depth to be credible and enough industry context to be relevant. When you can't find both in one person, consider which matters more for your audience.
A room full of engineers might prefer technical depth and can handle a speaker who doesn't know their industry. A room full of business leaders might get more from someone who understands their world and can translate AI implications into language they recognize.
Using Your Checklist Right
A structured evaluation process helps you compare candidates consistently and avoid getting swept up in particularly charismatic sales calls. We recommend building a simple scorecard that covers technical credibility, presentation ability, relevance to your audience, and alignment with your event's goals.
If you don't already have one, our booking checklist includes evaluation criteria you can adapt for your specific needs. The goal isn't bureaucratic box-checking. It's ensuring you ask the same important questions of every candidate so your comparison is fair.
What Credibility Looks Like in Practice
When you find a genuinely credible AI speaker, here's what you'll notice:
They're specific about what they know and what they don't. They'll say things like "I can speak authoritatively about deployment challenges but I'm not a research scientist" or "My experience is mostly in enterprise applications, so I'd frame consumer AI differently."
They have receipts. Not just claims about impressive projects but names, dates, outcomes, and ideally independent verification. They're happy to provide references from people who worked alongside them.
They have opinions that aren't just conventional wisdom with better presentation. They've thought hard enough about the material to have developed a perspective, and they can defend that perspective when challenged.
They're still learning. The best AI speakers are the ones who are genuinely curious, who read papers for fun, who are running experiments in their spare time, who can't stop talking about the latest development that surprised them.
They've been wrong publicly and can talk about it. Nobody has a perfect track record in a field this volatile. The experts who pretend otherwise are either lying or haven't stuck their necks out enough to be proven wrong.
The Investment in Proper Vetting
Thorough credential verification takes time. You might need to watch hours of video, make multiple phone calls, and do real research on someone's background. Event planners often feel they don't have time for this, especially when managing dozens of other details for a complex conference.
But consider the cost of getting it wrong. A bad keynote doesn't just waste the speaker fee. It wastes the attention of everyone in that room, damages your credibility as a curator, and potentially sends your audience home with bad information they'll act on.
The hour you spend verifying whether a speaker actually did what they claim is probably the highest-leverage hour you'll spend on speaker selection. It's certainly more valuable than another revision of the room layout.
We built Crimson Speakers specifically because this vetting is difficult and time-consuming. Every speaker in our bureau has been through a rigorous evaluation process, and we maintain relationships with them over time so we know when circumstances change. But whether you work with us or handle vetting internally, the principles remain the same: look past the marketing, verify the claims, and demand specifics.
Your audience deserves a speaker who's earned the stage. Finding that person takes work, but the work is worth it.