Well, good morning.
I would like to welcome everyone to today’s webinar brought to you by Atrium.
So, some housekeeping before we get started. First, this webinar is being recorded and it will be provided to you all by email once complete.
We will also have a question and answer segment at the end of today’s webinar. So if you have any questions regarding any of today’s content, please don’t hesitate to enter those into the chat.
And certainly, if we have time, we’ll, we’ll get to those. And, and behind the scenes, we’re, we’re gonna have John collect them and we’ll loop back to these at the end of today’s program or afterward. But let’s get into it.
So, AI in the workforce, is it a problem or is it an opportunity?
Like most things in life, the answer probably lies in between.
And so today’s topic is the AI hiring surge, what leaders need to know now.
And we are gonna cover why AI has become a workforce issue, not just a technology trend trend.
And so, why is this important?
So, it’s important because it not only changes how work gets done, but who does that work. And the entire workforce is collectively asking the same question, which is what specific AI skills are valuable.
So today I have the pleasure of being your host, and I am humbled to moderate alongside two incredibly talented and experienced individuals who I’m grateful to work alongside in our business every single day.
Um I am the, uh, Adam Samples. I’m the president of Atrium Talent Solutions, and I’ve had the absolute privilege of spending more than 20 years partnering with organizations all over the country and navigating their human capital strategies in the staffing category, and I’m, I’m grateful to be complimented by senior executive Vice President, Sara Mackessy and Vice President Marc Wilder. So, Sara, Marc, welcome to you both.
Thank you,
Adam, happy to be here.
Excellent. So, Sara, Sara has nearly 20 years of experience running both revenue and operations at Atrium in Markets like New York, Boston, Atlanta, and our Florida markets. And when she’s not partnering with clients and their talent strategies, she’s heavily involved in our business transformation efforts, including, of course, all things AI. So, again, welcome, Sara.
Marc has nearly 20 years as well of experience running revenue in London and Metro New York with really deep subject matter expertise in the IT segment.
Today he runs one of Atrium’s largest revenue streams, and he’s also celebrating his precious arsenal for winning their first English Premier League championship in 22 years. So welcome, Marc, and congrats. Thank you.
Collectively, we are going to be covering the AI hiring surge, and I thought a helpful place to start is to look back on the history of the last 4 industrial revolutions and really isolate why this one is different.
Most major technology revolutions have primarily focused on the automation of physical labor and enhancing infrastructure.
So whether it was the advent of steam power in the late 1700s, electricity and assembly lines of the late 1800s, or the computer age in the late 1900s, what these all had in common was that muscle was being replaced with machinery and routine tasks were being replaced by computing.
And of course, the labor shifted with it.
Today, companies across all industries and not just tech are being pushed into AI enabled roles.
Only this time, the impact is hitting knowledge workers.
Functions like writing, coding, analysis, research, and design are all being disrupted.
And put very simply, AI is materially impacting the very core activities of white collar work.
And so this leaves us with a central question.
In the race to becoming AI ready, what should leaders prioritize first?
Hiring new talent Reskilling existing teams, or redesigning roles altogether.
Let’s get started by simply talking about the market surge in AI talent demand. And so we’ll start with Sara here.
Sara, how has this demand surge changed in the last 12 to 18 months?
Thanks, Adam. Um, yeah, it’s a great question. So, let me start by giving you a number that I hope, that I think really helps to put it into perspective. In September of 2024, About 34% of tech job postings required AI skills, but by March of 2026, just earlier this year, it was 67%. So, that’s a 167% increase in just 18 months. And what’s interesting is we’re not just talking about dedicated AI engineering roles, we’re talking about project managers, analysts, customer service, marketing.
So AI fluency is really becoming a baseline expectation across the board, and I think what’s changed the game has been this shift from experiment experimentation to really production. Like a year ago, companies were piloting, they were experimenting, but now they’re really operationalizing, and once AI moves from really a proof of concept to core infrastructuring, hiring needs to change overnight, and that’s exactly what we’re seeing play out in real time, um, I think what’s interesting is it’s not just the obvious roles that are growing.
For example, LinkedIn called out that AI engineer is the fastest growing job category going into this year, which makes good sense, but if you think about the skills that are really exploding underneath that, their deployment, governance, their operational roles, and These are the rules that really keep AI running safely once it’s live, and it tells you where companies are right now, right? They’re past the let’s build something phase, and they’re, they need people who can really own it, who can manage it, and really answer for it.
All really good points and LinkedIn probably has the deepest pool of, of data that we can rely on here. So, we’re in early innings in this. There’s only gonna be more functional roles that pop up. But if we, if we talk about industries for a second, what would you say are the industries that are feeling the most pressure?
Yeah, that’s a great question. So, historically, you know, we’d say tech, finance, professional services, and yes, those industries show the highest concentration of AI skilled job postings, but I think the really interesting story is actually what’s happening outside of tech. So, in January of this year, insurance, for example, posted 145% month over month growth in tech hiring, which is the highest of any industry. Healthcare was up 72%, automotive was up 65%.
So, these are industries that are under urgent pressure because they’re regulated, they’re consumer facing, and they’re really competing for the same limited AI talent pool as every major tech company out there. Um, another, you know, example is healthcare and life sciences. are at a particular pressure point because they have to meet HIPAA, FDA, state level AI governance requirements. So they’re not just hiring AI talent, they’re hunting for people who understand AI within a really good compliance context and It’s just such a narrow pool, so the wait times are, are really reflecting it. Um, for example, McKinsey found that financial services and healthcare right now are waiting 6 to 7 months to fill a single AI role. Um, and I think a great real-time example, we’re seeing this play out right now, Cognizant is one of the world’s largest IT staffing firms.
And they just recently announced that they’re cutting 4000 roles, while at the same time planning to hire more than 20,000 people in 2026. So, if you think about that, their, their total headcount is actually gonna go up, but the composition of those hires is changing, right? They’re explicitly repositioning early talent, that is AI trained. So it’s not really a layoff story. It’s, it’s a workforce transformation story, and it’s happening across every industry, not just in tech.
Wow, wow. So it’s just industry agnostic, um, not surprising, and it sounds like even, even the more historically slow to adopt technology industries are, are all over this. So, I guess from a competition standpoint, how competitive has the, has the AI hiring become through all these industries?
Very. So, extremely competitive. I, I would describe it as structurally competitive, meaning it’s not gonna resolve itself in the next hiring cycle. You know, an interesting stat to think about is globally, there are 1.6 million open AI roles, but there are only 500,000 qualified candidates. So, That’s leaving us with essentially a 3 to 1 demand to supply gap, right? And um the other, the other piece that I think is important to think about is comp, right? And that’s where it starts to get really hard for most organizations. Um, for example, PWC tracked The wage premium for AI skills over the past year has doubled.
So, if you think about that gap between what a Google or an OpenAI can put on the table versus, you know, what a regional bank or a healthcare system can offer, it’s just not close. So, when you’re a mid-market employer, and you can’t just outspend your way to AI talent, you need to get creative.
Um, and I think what makes this genuinely hard for hiring leaders is that You know, a hospital system, a retail bank, even a Fortune 500 CPG company, they’re competing for the same prompt engineers, the same LLM ops talent, the same AI compliance specialists that Google and OpenAI are hunting right now. So that talent pool is all the same, but the budgets aren’t. So, There is this central tension that we need to help our clients really navigate right now, um, and I think that the companies that, you know, the companies can’t win on comp alone, those companies are just starting to get creative about their org design. So another recent example actually just happened a couple weeks ago is Coinbase cut 14% of its workforce.
And they announced that they’re rebuilding what they’re calling these AI native pods, so it’s essentially one person teams where a single employee supported by these AI agents cover what used to require an engineer, a designer, a product manager. So it’s not really a hiring strategy, it’s, it’s a reimagination of how to get work done, and it’s, I think, a signal of where most competitive organizations are, are heading.
Well, that, that 3 to 1.
The gap is compelling. Uh, it sounds like yet another talent war and that’ll be stacked already on the 4.3% unemployment rate. We’ve seen this, we’ve seen this movie before, but Marc, I’d love you to chime in here. Any, any additional thoughts on this?
Yeah, I would just add, um, you know, Sara said that any market, be it hiring or any other, is dictated by supply and demand, right? And um we know from those stats that Sara gave us how high demand is for AI talent, but there’s a supply factor here too that I wanted to bring up. Um, traditionally, a primary pathway for foreign workers to enter the US uh has been the H-1B visa application.
And due to new, um, legislation from this administration. Which places a much larger financial burden on organizations to um go through the H-1 visa visa application process. We’ve seen H-1B visa applications drop dramatically. So it’s predicted this year there will be a 24, 25% drop from last year to this in how many H-1B visa applications are put in, um, and that’s a 4 year low. And so it’s just exacerbating the issue further that you’ve got all this demand and even less supply than normal.
Just another pylon.
Really good point. Well, let’s stick with you, Marc, here. So if, if we talk about the new AI job ecosystem, what, what would you say the new AI roles are that organizations are hiring today?
Yeah, good question, well, you asked about new roles, so let’s, for the sake of going through this and dissecting it, let’s put to one side AI engineers, machine learning engineers, and assume that everybody on this call have heard of those already. Um, we’ll start at the very top of an organization and a role. That we’re seeing emerge now and will continue to do so is the Chief AI Officer, the CAIO and just like the chief level in any capacity, Chief Financial Officer, this person is responsible for overall enterprise wide strategy.
Um, for some context here, a recent IBM survey said that 66% of large organizations are planning on hiring a chief AI officer if they haven’t done so already. And so this is gonna be widespread, you’re gonna be seeing this a lot. Chief AI officer at the top of the organization again, are gonna cover AI strategy for a, for a company in totality.
Other roles to explore, a prompt engineer. So, you know, you or I using chat GBT, the things that we’re entering to an AI system are known as prompts, of course, uh, a prompt engineer, this is a role that really, I mean not heard of at all before say 2023 or 2024, but. It becoming a big area, this person will be responsible for precise inputs to AI. Um, if we want accurate and useful outputs, somebody needs to be putting in the correct prompts, and organizations are employing the prompt engineer to cover this space.
Level above a prompt engineer as a context engineer, this is somebody that ultimately is gonna design the systems that give AI the right information at the right time. You know, AI works on models, and those map models are only effective if they are grounded in accurate data. And so um context engineer, providing the context in which AI is working, context engineer, it’s another big role for the future.
Some other ones perhaps are less well known, AI agent architect.
You know, somebody needs to be responsible for designing the systems where AI models are gonna interact with say a company’s tools or a company’s data. We want AI to work autonomously, um, that’s where the magic happens, right? AI working autonomously with data and tools that already exist in an organization, but somebody needs to design that system where AI interacts with tools and data, so the AI agent architect is gonna be responsible for the, for that.
AI is best and most effective when working with human beings, and so, um, a title you’ll see more and more is the AI human workflow specialist. This can be called a couple of different things, but AI AI to human workflow specialist is gonna design specifically how are people collaborating with AI. If we wanna maximize efficiency, um, somebody needs to be responsible for that workflow. When, how, where to use AI within a company.
Um, AI trainer is another one, and contrary to popular belief, this isn’t just somebody coming in and training everybody how to use AI. AI itself needs training to be most effective. And so being responsible for curating the data and crucially the feedback that’s gonna teach the models to become more accurate and more reliable over time, that is an entire role in itself, so AI trainer.
Um, we already went through a lot here, but one more to leave you with is, um, probably the biggest area I haven’t mentioned yet, AI compliance and governance.
You know, if you think about companies are implementing AI, if you don’t have the right guardrails, risk management, regulation, um, and really like responsible use within an organization, being quicker by using AI turns into being. Exposed pretty quickly from a governance and compliance perspective. So it’s already happening, but expect AI compliance and AI governance to be a huge growth area, especially in highly regulated environments, so think financial services and healthcare.
Compliance and governance, those are two words that captures everybody’s attention and for all the wrong reasons, right? So, um, yeah, that’s really, really helpful, Marc. And I think, you know, even a year ago, while we knew these were all buzzwords, none of these roles, uh, is anything we’ve ever heard of before. Why is, why are these roles emerging so quickly?
Well, similarly to why so many people are tuning in today, um, you can’t turn on the TV or be in a business meeting or uh listen to a commencement speech at a university without hearing AI being spoken about. It is, uh, without question the thing that is on everybody’s mind.
And you’ve gotta think contextually about the level of investment that’s actually going into this space right now. Um, so I saw that if you combine Amazon, Alphabet, Microsoft and Meta, just four companies, and you analyze how much they actually are committed to spending on AI in 2026 alone, um, it’s about $700 billion which is like an astronomical figure to even get your head around.
For context, that’s roughly what the US government spends on Medicare.
Um, and you know, you hear about Mark Zuckerberg’s reportedly offering like ridiculous packages to lure AI talent to meta, $100 million 300 million dollars dollar package ranges, even a rumor about him maybe offering a billion dollars to somebody.
Um, you know, just the level of investment is truly unprecedented, and that’s why roles are emerging so quickly. And I’ll just add, you know.
A lot of companies report that it’s not, it’s, it’s an operational issue as much as it is a technical one. Meaning, you know, these roles are emerging to provide strong and safe execution and fix failures in AI to keep up with this, um, you know, 600% increase in AI job listings that have gone up in the last 3 years.
Wild numbers, truly.
And being that they’re so new, and there’s such a high price tag coming with these, how important is real domain expertise in these new AI roles?
I would just say Today in 26, um, specialists are more sought after than generalists, um, when it comes to AI hiring.
Deep knowledge in say one aspect of AI, um, is actually more valuable than say, shallow knowledge across a lot of areas. You know, there’s that phrase that’s like an inch wider and a mile deep or, or whatever that phrase is, um, you know, that, that really applies here in AI, um. We’re really seeing the most valuable hires, the ones that can like can kind of um combine, let’s say AI fluency with really deep knowledge of a specific industry or a specific function. So yes, I have AI skills.
Put it this way, a machine learning engineer who understands how a CFO thinks, who understands the challenges that a finance organization has, is more valuable than one that doesn’t, and we’re really seeing that play out in the staffing world like every single day.
Yeah, that makes a lot of sense.
Sara, any, any additional thoughts on this topic?
Yeah, I, I think exactly like Marc said, we’re seeing this play out constantly. Again, the cognizant story I think is actually a good example of this, like they’re not just hiring. AI talent broadly, they’re specifically repositioning towards people who understand their delivery model and their clients’ industries. So I think ultimately domain context isn’t really a nice to have anymore. It’s ultimately becoming kind of the, the differentiator between a hire that works and one that ultimately doesn’t.
Yeah, we’ll put.
Well, if we take the functional roles we we’re talking about, and we, we, we pass that into the skills that go along with those, and just talk about the, the growing AI skills gap for a second. Sara, we, we can start with you. What, what is really driving the AI skills gap right now?
Yeah, it’s an interesting question. So I I would say the gap has two main drivers, and they’re both accelerating at the same time, which makes it particularly hard to close this gap, but the first that might be obvious is just supply, right? Universities and training programs just simply cannot produce credentialed AI talent fast enough, so, I mentioned the stat earlier, but we currently have 1.6 million open AI roles globally against roughly 500,000 qualified candidates. So, it’s just not a gap that gets solved in a single hiring cycle, and I would say actually by most projections, it’s gonna get worse before it gets better. Um, the second, which we’ve spent some time talking about already, but it’s important is compensation, really velocity. I mentioned.
Again, this stat also, but PWC found that the wage premium for AI skills doubled in a single year. So, even the organizations that find candidates, they’re struggling to close offers because most organizations just can’t write the kind of offer letters that these big tech firms can. So, I mean, that’s just the reality right now. So then the question becomes more how do you compete differently. Um, I thought this was an interesting, um, stat from the World Economic Forum, but they put it pretty starkly, 63% of global employers cite the skills gap as Their primary barrier to business transformation. So, the irony there is that organizations that do have AI capabilities are pulling ahead so fast that that gap between the AI enabled companies and everyone else is just widening and widening by the corridor, so it’s creating this urgency that we really haven’t seen in the workforce planning in a really long time.
Yeah, the 3 to 1 gap, second time it’s hit. So, when we talk about the the struggles organizations are facing, how, how are they struggling most with just AI capability in general?
Yeah, so, I think there are two places where the gap feels most painful right now, and interestingly, neither of them are pure engineering. I think the first, Marc had mentioned it earlier, but Governance and compliance, and as AI gets deployed at scale, someone has to own the risk, someone has to own the bias auditing, the regulatory compliance, and these are brand new roles like Marc mentioned earlier that didn’t have a career path 18 months ago, right? Responsible AI. Cybersecurity compliance, they are amongst the fastest growing AI adjacent skill categories this year, and I think companies are just scrambling to staff them because they moved fast on the deployment and then they realized they needed the oversight infrastructure.
Um, the second is what I call this translation layer, so the people who sit between the business and the technology and then actually connect the two. So, someone who can take a messy operational problem, and then turn it into something AI can solve, and that role sounds simple, but it’s genuinely rare, and I think the organizations that have it are the ones who are actually starting to see ROI on their AI investments, and everyone else is still waiting for the technology to really just prove itself.
Uh, classic order of operations challenge.
So, when we think about downstream effects from these gaps, what, what would you have to say there?
So, there are a couple of downstream effects that are hitting, um, I think primarily in let’s 3 different ways and each one really has real dollar consequences. So, the first is purely time to fill finance and healthcare, we had talked about um earlier how long it was taking to fill jobs, but they’re averaging 6 to 7 months to fill a single AI role. So, When you’re trying to launch an AI-driven product or comply with a new regulation, right, a seven-month hiring cycle is an existential risk. It’s not just this inconvenience that’s sitting in an organization.
And then the second, which again we’ve, we’ve spent some time talking about compensation, but it’s, it’s very real, compensation inflation, and that data is very clear. So, companies that waited 6 months in early 2024 to pull the trigger on AI hires ended up now paying a 15 to 20% premium for that same talent.
So, I think the messaging there is just every quarter you delay, the market keeps moving and it’s just, you know, there’s obvious cost there of, of inaction, ultimately. And then third, and I think this one is probably most underappreciated, but competitive distance that’s, that’s happening right now. So companies with established AI teams now have a 22 month lead over competitors that are still building their capabilities, and that lead just compounds and it compounds. So AI systems That have been in production for 2 years are gonna generate better data, better training loops, better outputs, so that gap between organizations that moved early and those that are catching up isn’t just headcount, isn’t just a headcount gap, right? It’s, it’s a capability gap that, that will take years to potentially close.
And then lastly, I would say, kind of this concept of organizational whiplash happening. So we’re already seeing some companies that made aggressive AI-driven cuts and then turn right around and rehire those same roles months later because the strategy moved faster than the thinking did. So, about a third of these organizations that did AI related layoffs had backfilled up to half those roles within 6 months, and that is not only expensive, but The hidden cost there is what does it do to the people who stayed, right? You, you, it’s very hard to build an AI ready workforce force on a foundation of just broken trust essentially.
Certainly is, that’s disruptive to say the least.
Marc, anything to add to this topic?
Look, clearly the, there is a real, very real issue about a skills gap in the marketplace today when it comes to AI um, the number one workaround that we’re seeing from hiring managers and leaders of teams out there today is um getting around some of these challenges, looking at working with specialized contractors. So, uh, this is especially true, by the way, when you talk about like a defined phase of work.
Rather than just constantly trying to hire full-time talent that doesn’t exist.
A lot of people are pivoting, and I saw this stat earlier, 92% of tech leaders plan to increase freelancer engagements in the next two years. So it’s not just that they’re doing it already, it’s that they see the future and realize that this is probably the only way that they’re gonna be able to secure the talent that they need.
Yeah, the fastest way in in a a classic way to fill those gaps. So let’s stick with the The skills gap for a second, really start to talk about why upskilling your existing workforce starts to matter right now. So, so Marc, for you, we’ll stay with you. Why, why is upskilling so important for AI readiness?
Well, it starts with everything we’ve been talking about, you know, um, external hiring alone just is not gonna meet the demand, right? If the people aren’t out there, there’s only so many people you’re gonna be able to hire in those spaces. Um, we talked about, um, the World Economic Forum have released the Future of Jobs, uh, report that says 80% of the global workforce will need to acquire new skills by 2027.
You think about it from a company’s point of view, Adam.
A company would prefer internal mobility if it’s possible. It’s faster to do that. You know, Sara was talking about 6 or 7 months to hire somebody in some spaces, so it’s faster to hire internal, um, to upskill internal talent. Um, it’s cheaper and it’s easier if it’s done correctly. Um, from a, from an individual person’s perspective, you can think about it like.
Proactively upskilling is just making you proportionately, disproportionately more valuable, um.
Another way to think about it is like everyone does not need to become an AI engineer, that’s, that’s not what I’m trying to say. It’s more like you do need to be the person on your team who understands how do we use AI to produce results within our specific domain.
And you know, it’s come up already, this, this story with Cognizant about laying off 1% of their global workforce, um, 4000 employees and hiring back, um, 20,000 more who are AI trained talent. That’s gonna save them $200 to $300 million a year, but the message I read out of something like that, and I don’t mean to be too much of an alarmist here, but essentially it says to me.
Gain AI skills or the risk of being replaced by somebody that has them is increasing more over time, not less.
Yeah, makes sense and we’ve, we’ve heard that in, in a lot of different phrases, uh, for really the last year or two, and it’s something that people are talking a lot about, but, you know, it does make sense to use your existing workforce, retrain them.
That institutional knowledge can be very valuable if, if, if you can retain those and, and really redeploy. It’s, it’s an important Revolution. So, if we, if we get into the specific roles, um, that are best suited for AI upskilling, what, what pathways do you see there?
Well, it’s almost easier to talk about roles that aren’t um easily upskilled by AI than are because it really, I mean it affects almost everything. Um, let’s start with our world, so talent and recruiting, I lead talent and recruiters, um, sales uh people and, and recruiters here at Atrium, um.
We don’t want to remove the human element to these roles, and you know, that is such an important part. When we define what a recruiter or a salesperson at Atrium does, it’s um part art, part science, and the art part is that human connection, building trust with candidates and clients. We don’t want to upskill that area with AI.
It’s the science part. So when you think about sourcing resumes quicker, um, outreach or market mapping, the idea is that we keep the human connection, but AI can 10x your productivity without losing the art piece of the role. The idea is, you know, our people will get sharper and faster but not replaced.
One of the biggest areas, you know, obviously leading a tech team, software development has to be mentioned here, um, AI has had just an enormous impact on software development.
When you think about, you know, everyone right now is talking about Claude code. Adam uh himself recently just built an app that we’re using inside of Atrium today. um, Cloud Code and other tools like this have turned everyone into a software developer, uh like overnight.
So, um, what we’re finding in the software develop, the real software development world is that AI is handling like a lot of the routine coding. Um, where we’re finding a real shift is the premium that’s being placed now on the ability to understand and solve real business problems by using software, using coding. So that’s how the software development world is kind of being upskilled.
Um, it was mentioned earlier before, Adam, I think you talked about analysis, right, finance, operations, anything involving data essentially, just the time to insight, right, I have my information, how quickly can I get something that gives me real information as to what’s going on. That time has obviously been just dramatically um reduced.
Um, our temp admin team here at Atrium does a lot in the customer success and customer support world. Um, AI’s handling volume and knowledge retrieval quicker than ever. Uh, humans are there to focus on the relationships, as I kind of mentioned with staffing before.
Um, marketing and content, so areas you can think about like copywriting, SEO, analytics, AI’s accelerating the output. Humans are there to keep thinking about branding and strategy and keeping that tight.
Um, and then HR and operations is another area I mentioned, be it policy drafting, um, onboarding, uh, performance frameworks, internal communications, AI is speeding the administrative part of the roles, and humans inevitably are handling the nuance part.
Well, Hopefully, we’re all in for a lot more capacity to just be more productive and, and do what’s being regarded as more human things while we automate the things that, that have taken us so much time for so many years. So, if, if we think about in order to pass to what I just described, people need to know what they’re doing, right? So, when it comes to just really effective AI upskilling, what, what does that look like in practice?
I think probably, probably two points to make here, um.
AI upskilling works when it changes how people actually do their job. What it’s not is sending people on a course and they come back and they know a bit more, but they work essentially in exactly the same way. It’s really, it’s, if it’s changing the way people do their jobs, that’s effective. Um, and again, you know, we’ve talked about domain experience a bunch, gene generic training is just less effective.
Um, really it’s, it’s the where we’re seeing multipliers is existing domain knowledge and human skills interacting together, um, with the upskilling in AI. And then the second point, I know no doubt there’s a lot of, um, learning and development and HR leaders that are on this call. Um, this, these functions can really add very real value here. Uh, you need to be designing and scaling role-based training programs. Um, either way.
I would, I would just say the time to do this is right now. Rather than just purely trying to go for perfection, time is of the essence. And, you know, an interesting stat that really hit home for me was that 68% of corporate employees today are using AI in some capacity in the workplace without formal guidance and structured programs. And so, you know, sooner rather than later, we need to be rolling out very specific AI training programs to ensure things don’t kind of get out of control there.
For, for a litany of reasons, I’m sure, but right, I think we all know people have gone rogue and they’ve started to become curious and, and, uh, that needs some guardrails, but it also needs some, some training to make sure that they’re using it efficiently. I get it.
I get it. Well, Sara, talk to us. Anything to, to add to what, uh, Marc shared?
Yes, I think, well, first of all, that stat is striking, 68%, and it I think it means the upskilling is already happening. It’s just informally and inconsistently happening without any sort of quality control, so the organizations that get ahead of that with structured programs aren’t starting from zero, they’re, they’re making really intentional what’s already happening organically, so I think that’s actually a much, much easier lift than most leaders think, and like Marc said, I think the time to act is now.
Indeed, indeed, well put.
Well, we’ve talked so much about Different channels by which we, we go about this, this AI readiness strategy and, and of course, our shameless plug, we want to talk a little bit about how staffing partners can add value in this equation and historically, uh, whether it’s bringing in full-time talent or, or, or interim talent.
I think we want to focus on two questions here. So just a very generic, where do staffing partners add the most value in AI hiring and workforce strategy.
But also how organizations should think about staffing partners as a part of their broader approach to their AI strategy. So, I think to start like in general, it’s really about acting as a strategic partner, as a staffing agency to really help clients architect a blended AI talent acquisition approach versus simply filling open requisitions, which we can do. But that would mean leveraging the constant market intelligence gathered through experience with hundreds of other active clients and doing that in a few key ways.
First, really helping to audit.
What capabilities any given client has versus what they need.
It also means advising on which roles to hire permanently, which hires to, to bring in on a contract basis, and maybe which roles to just deprioritize altogether.
Um, it means sourcing across a lot of different talent tiers to get access to emerging and niche AI skill sets and not just going to that first tier that everybody is very comfortable with.
And it means educating, educating clients on realistic market expectations, for example, for compensation, for timelines, and, and for specific talent profiles, um, versus falling into the pit of trying to hire an all in one generalist.
All of this will help drive Really faster deployment in what are clearly undefined and very rapidly evolving role categories. And, and so by operating as extensions of customers’ internal talent teams, staffing partners bring market insight and, and flexibility.
That can empower clients to move faster, to benchmark more effectively, and also to adapt as, as needs shift. And we all know that they will indeed shift. I feel like they’ve shifted 10 times just in the last 3 to 5 years. And, and as they shift faster and more often than most people are, are comfortable with, and, and we don’t see any, any end to that dynamic. And so, in summary, look, We have covered a lot today, a lot of heavy content, you know, from the AI talent demand to the AI skills gap.
We touched on the functional roles that are impacted on one end of the spectrum, and we talked about the variable compensation for these new functional roles created on the other end of the spectrum. So, uh, before we get into some of the Q&A, Sara and Marc, any, any closing thoughts to what we’ve covered today?
Sure, yeah, I’ll jump in. I think It’s interesting because this webinar is obviously focused on AI, but ultimately, I think everything that we’ve talked about today goes back to people, the data, the job titles, the skill gaps, they’re all systems of the same underlying reality, which is that this transitioning is happening fast. And humans are navigating it and they need real support. So, the organizations that get that right, the ones that invest in their people’s ability to really evolve alongside AI and not just hire around them, those are the ones that are gonna come out of this with both competitive advantage and also I think just a workforce that actually trusts them and that feels to me at least like a win that’s worth building towards.
We’ll put Marc.
Yeah, I’ll echo some of what Sara says, um, this is changing month to month, literally.
Um, AI really can be thought of, this generation’s industrial revolution, but this one is happening at the speed of light. And the level of investment that’s going into the AI space is unprecedented.
Um, like I said before, what we’re finding is companies are using a wide range of options to support AI initiatives. It’s really the, the ones that are doing this best, cos it is challenging for everyone, but the ones that are doing this best are doing so with a multi-pronged approach, it’s not one thing to solve the challenges of AI.
It’s upskilling your current staff for sure. It’s hiring strong talent from outside, but also you can’t be afraid to supplement, especially for defined work, contract or contingent staff.
Um, what absolutely hasn’t changed is trust, and I don’t think it ever will. Sara touched on it. Um, I would advise everybody on this call as they think about navigating these choppy and difficult AI hiring waters, um, build trusting partnerships with talent organizations that can support you, um, by doing not one thing well, but lots of different ones.
Uh, really well put, and what it is being repeated over and over again are people in this equation, which, uh, which is ironic considering the topic, but, you know, the AI romantic in me truly does believe that by automating so many tasks that have been so manual for so long, it’s really gonna free up human capital to do the things that humans were.
We’re born to do. Um, and so, look, we talked about key takeaways for a moment before we get into Q&A and I, we really want to leave you with two key takeaways. And I, I’ll start with a thought provoking question. So, why should AI hiring be treated like an emerging market rather than a standard recruiting cycle, right? So, first, let’s, let’s start with the talent supply in this category. The reality is, these talent pools in, in AI right now, they’re nascent.
And true AI practitioners, those who can build, fine-tune, deploy, and govern AI systems, are still a very small and very concentrated population. And just like in an emerging market.
You can’t always apply the same tried and true methods of sourcing talent that have worked in the past. Talent is often not found on the same platforms or even in the same pipelines, uh, or responsive to the same type of recruiting outreach that we all know so well. And so the reality is that roles, skills, and compensation.
are shifting faster than traditional workforce planning models. And if we first talk about the roles themselves, they simply aren’t cleanly defined as we established earlier in, in, in this webinar. And because there is no deep historical data to draw from, hiring often requires building specs from scratch. Um, there is no playbook.
And so if we talk about skills and education, these degree programs in AI proper are only just starting to emerge and not yet widely distributed. And, and finally, you know, on the compensation front, AI compensation is volatile, and it’s not yet anchored to what most HR departments are conditioned to establishing.
Which is these very organized salary bands. And, and so, you know, I think the second takeaway really addresses the very essence of today’s topic, which is what do leaders need to know now when it comes to the AI hiring search. And, and so that begs the question, what does an effective AI talent strategy blend look like?
And when it comes to an effective framework to apply to your AI talent strategy blend, we want to leave you with with the three B’s build, buy, borrow, and bought.
And so let’s start with build. What do we mean by build? Well we mean internal upskilling to scale capability quickly and cost effectively. And Marc touched a lot on that earlier, and this is best used when you have domain experts. So that’s that those domain experts with their institutional knowledge, bring them to life by upskilling them with AI.
Outside of build, we have buy, and buying, as you, as you can imagine, would be going to the external market and hiring for very scarce or, or leadership level AI expertise. And, and so when deep technical ability is needed that can’t be developed fast enough, buying is your strategy.
Move on to the third, we talk about borrowing.
And and this is where the contingent labor strategy comes into the fray. Contingent talent is used often to address short-term gaps and to do that very quickly. And so when you’ve got rapid change, think about bringing contingent talent into your ecosystem to, to get lift off the ground and have these people hit the ground running.
Uh, when your needs are specialized, when your needs are time-bound, or if they’re simply just experimental and exploratory, and, and the fourth is kind of the newer one, right? That’s bot, right? Bot it. You’ve got to automate, automate tasks with AI tools rather than adding to headcount is a 4th, uh, strategy to be used here, when the work itself can be transformed and not just the person. And so, in closing, you’re gonna need AI builders who can create.
You’re going to need AI integrators who can cross-functionally deploy your AI and you’re gonna need AI enablers, and, and this is where blend and balance really matters. And so, we are coming up on a pretty good timeline here. We’re coming at just about time, but we, it looks like we do have some questions that have been submitted. So, John, why don’t you tell us what the audience is curious about on the AI front?
Yeah, Adam, we got a couple questions and one of the first ones that came in was, what’s the biggest mistake companies make when designing AI job descriptions?
OK.
What’s the biggest mistake companies make when designing AI job descriptions?
Marc, you wanna take this?
I would say not dissimilar to any other um areas, but um I think it’s like the thought process behind why do we need this job.
Um, what I’m talking about here is if companies start with I need an AI engineer, rather than we need a problem solved, I think that’s where you can come unstuck, right? There just needs to be a little bit of more thought that’s put in for a couple of different reasons. One, When you start with I need an AI person to come join our company.
The job description that’s being written ends up being it’s buzzwords, right? And it doesn’t describe a problem that we’re trying to solve, and I think talent at this point is already savvy to some of those things, and I think weary of a very generic job spec or, or job details. I think job descriptions need to be specific to align with specific talent that exists in the marketplace. And so. You know, another rule of thumb I think that maybe hiring managers or leaders who want to hire in the AI space can be thinking about is if you can’t articulate why you need this person, or more specifically what this person will change in your organization or your group in the next 12 months, there’s a really good chance that you’re not quite ready to go to market and find somebody yet.
Um, what you actually need to figure out is that one step before. What are the problems that we’re trying to solve, let’s put that clearly in the job description, so we can attract talent that matches that specifically, otherwise generic titles, you know, are not going to produce the best results. I think that is probably the, the biggest problem that we’re seeing today.
That’s really well put. So, so really describing.
What they wish existed is not the right play. It’s really being very specific about the work that actually needs to be done, right? So, you also touched on that, that, that pit that everybody falls in of, of, I think what’s known as the everything AI, right? I just need, I need it all, right? So really getting narrow about what the, what the problem to solve is.
Very good.
What else we got, John?
Another question is, um, what AI governance roles are companies underestimating right now?
OK.
Well, I’ll, I can just touch on that for 2 seconds. I think what, what’s more important to highlight here is the number of companies who consider AI governance as a compliance checkbox that, you know, your legal department handles or, or really just something that exists on a, on a, on a document somewhere and that function really must be uh considered uh simply managing risk for the systems that are making decisions at scale. This, this is requires human expertise.
And that doesn’t exist in the majority of today’s org charts, which will come as no surprise to anyone. So, we’re talking about roles such as AI risk officer, um, AI auditor, AI policy and regulatory affairs specialist. All, all roles that, that, that sound similar to this are, are what’s being underestimated right now are not even being thought of altogether.
How are you doing,
Adam?
Yeah, we had one more if you think we have some time.
Yeah,
I think we got one more.
Sure, great. Um, the one other question is, How realistic is it to move existing employees into these AI roles?
I can take, I can jump in here and take this.
So I actually think it’s far more realistic than most companies think, but harder than the training programs will tell you, right? So I think again the truth is probably in the middle and it comes down to the role ultimately, but the transitions that work are the ones where the people already have that deep domain context which we’ve talked about today. So your analysts, your recruiter, your ops manager, they don’t need to become AI engineers and they, they frankly shouldn’t be expected to, but they do need to become the person on their team who knows how to get AI to produce results in a very specific workflow, and I think that when you break it down like that, it’s, it’s achievable.
So, where it ends up breaking down though is when companies try to shortcut it and then You know, they decide that upskilling just doesn’t work, but ultimately it was when it wasn’t the right person, it was the wrong role, it was the wrong training design. So the jobs ultimately aren’t, aren’t being replaced here, they’re just being redesigned, and the question isn’t, can this person become an AI expert, it’s what does their role look like with AI as a collaborator, and I think we should just, my recommendation would be to start, start there.
Yeah, really well put.
We all just need to put on our AI armor, right? That’s, it’s just another tool in the tool belt, but it’s an important one. So, very good. Well, I think we’re, we’re close to the end of time here discussing, uh, this all important topic, and, and I hope you all have found value in the dialogue that we’ve delivered today. So, hey, Sara and Marc, I wanna thank you both for your brains and for your time that’s been devoted to this topic.
Um, And all today’s attendees and others who will listen on replay, thank you for joining us here at Atrium to discuss such an exciting and perhaps equals, uh, equally unsettling topic. I hope that you all walk away with a feeling of urgency, but urgency without panic when it comes to your own AI talent strategy. And, and again, just as a reminder to everybody, you’ll all be receiving an email with the recording for today’s webinar. Uh, and we have a survey that should pop up here on your screen.
There it is. Now, to uh share your feedback on your experience with us that that we can be, you know, used obviously for future.
Webinars, and of course, feel free to reach out to us directly here at Atrium should you have any questions or curiosities on what we are experiencing in the marketplace with, with the hundreds of customers that we’re servicing every single day. We would love to have those conversations with you and add any value in any way possible, sharing, sharing that information. And so, again, thank you all for joining and have a great day.
Thank you.
