PwC Coaches New Tax Associates on AI Tools in US Training Effort

PwC’s new training program aims to give early-career recruits hands-on experience integrating artificial intelligence tools into everyday work.

_{Listen here and subscribe to Talking Tax on Apple Podcasts, Spotify, Megaphone, or Audible.}

The Big Four accounting and advisory firm started piloting AI immersion sessions in October, with a full rollout to new US associates slated for July. The sessions are happening across PwC’s tax, assurance, and advisory business.

“We truly believe that the role of the new associate will be changing with AI and that their role will become somewhat elevated, and we need to make sure that we’re really training them on those skills to work and think differently,” said Margaret Burke, the firmwide talent acquisition and development leader for PwC US.

Like its competitors, PwC has recently funneled resources into next generation autonomous tools aimed at handling routine tasks solo.

The firm said in November it shed about 150 jobs across marketing, human resources, and other US support roles as part of a longer-term effort modernizing its back-office unit, including through using new AI tools.

In this week’s Talking Tax, Burke and PwC US Tax Leader Krishnan Chandrasekhar sat down with Bloomberg Tax reporter Jorja Siemons to discuss how the AI trainings have gone so far and what skills they hope new employees learn.

Do you have feedback on this episode of Talking Tax? Give us a call and leave a voicemail at 703-341-3690.

This transcript was produced by Bloomberg Law Automation.

From Washington, I’m David Schultz, and this is Talking Tax.

AI is transforming the work of early-career employees, and companies are changing their onboarding programs accordingly. PwC recently started piloting AI training sessions with new associates across US teams. They’re trying to give younger workers hands-on experience using AI tools that are redefining daily tasks and client work.

PwC’s U.S. tax leader Krishnan Chandrasekhar and Firmwide Talent Acquisition and Development leader Margaret Burke have helped to shape these onboarding sessions. They spoke to Bloomberg Tax Reporter Jorja Siemons ahead of the company’s plans to extend this pilot program company-wide next summer. And Margaret started off by talking about how associates are reacting to these training sessions.

Margaret Burke:

I think we signaled to them right away that we were doing something different. We started right off the bat the day they arrived in the evening with an immersive hands-on keyboard activity that was built all around a theme park itinerary, and it was a fun food hack that they had to solve and bringing using social media, etc. And we did that to really signal to them right away that this was going to be different. And AI actually judged their submissions before they went to dinner. So I think it just showed them right away that this was going to be different.

We truly believe that the role of the new associate will be changing with AI and that their role will become somewhat elevated. And we need to make sure that we’re really training them on those skills to work and think differently.

Jorja Siemons (Bloomberg Tax Reporter):

The firm could have pursued other training strategies like asynchronous programs or virtual learning modules. Why did PwC opt for this specific way of teaching new associates about AI?

Margaret Burke:

I think, A, because it’s new, it’s new to us, it’s new to our associates. And like I said, our associates are the population that we felt as though was going to be affected the most immediately. And we also had the advantage of our new associates, they have a lot of existing training when they begin. So in some instances, we were able to add it on with tax. We brought them together separately just for this program. But I think we wanted to see how they reacted to it. We wanted them in a room together. We wanted them problem solving. We wanted to see, you know, their reaction to the tools, their ability to catch on to the tools. And listen, we had a lot of people coming in that were very comfortable with the tools versus those that weren’t. So how do we kind of blend that moving forward and make this work for them going forward?

Krishnan Chandrasekhar:

Yeah, one other thing I’d mention is everything Margaret described makes, you know, was spot on from a learning perspective. From a business perspective, very early on, we want to give them the intuition and the muscle memory that a lot of this work, we truly believe isn’t just about the tools. It’s about how you collaborate and cooperate in engaging with the tools. And we want to start that early.

So bringing them in and putting them in a room, as opposed to, we could easily do a virtual session, you know like, hey, AI is digital, let’s do it virtual. We actually want to change the muscle memory of our staff to say like, actually, you really get powerful in solving these issues when you are in a room with five other people and each one brings a different perspective, a different mindset. And collectively, when you see that with the tools, you’ll hear a lot from Margaret and me on the human side. That is an important part of what we want to emphasize in this training. It’s not one or the other. It’s all of it together is what creates the impact. So I just want to bring that up, too.

Margaret Burke:

Yeah, I couldn’t agree more, Krish. And I think really drawing that connection right from the AI tools to what they’re actually going to be doing in tax, right, was super, super important and to show them that the same skills are needed around empathy, around creativity, around human judgment, around critical thinking, all super, super important.

Jorja Siemons:

I want to turn to the tax business side specifically, as you both have mentioned. Krish, what are some examples of tax specific AI skills that associates are learning in these trainings?

Krishnan Chandrasekhar:

So in early days, and I’ll go beyond associates now a little bit. So what we just described as kind of new associates coming in today, where we are, is for each of the layers of our organization, associates, seniors, managers, I’ll stop there for now. What we are doing is saying, what are your current activities look like on a day-to-day basis for each of the levels and where are places where you can now team up with different sorts of AI tools at the right point in your current processes so you can enhance or have a thinking partner or an accelerator?

So if you are a staff that’s involved in review and analysis, what are new AI tools you can use? The first review you hand off to an AI language model, see what comes in and then have a Q&A to say, OK, what has it gotten right? What can I learn from this? What can I teach the next layer? And then deliver it. So I save some time by using the AI tool, but then use the extra time that I’ve created to actually learn a little bit, teach a little bit.

What I see happening, by the way, just to give a teaser and I’ll stop, is as we are doing this and also separately transforming some of our processes, we think fundamentally it might even change and offer opportunities for staff to come up with new ways of doing what we do today. We’re not there yet, but we see that be as the next thing we get to. But just starting to get them comfortable right now with current processes and enhancement and teaming, that’s what that’s what we’re zoned in on.

Jorja Siemons:

Can you give an example, Krish, of a lesson plan or activity in these trainings that teach some of these skills to tax associates that you’ve mentioned?

Krishnan Chandrasekhar:

Yeah, I mean, I’ll speak more on kind of how we are doing it for the layers of our organization that are already in. And then I’ll defer to Margaret on kind of lesson plans for the new associates. But think about for staff that are already in our hallways, the lesson plan is really current execution process. So we’re not necessarily coming up with a plan that says, OK, here’s the steps and here’s all the places you have to insert. We’re kind of letting them experiment a little bit and say, here’s all the steps. Here’s how teammates have surfaced either customized GPT models that allow you to quickly query things.

Like state and local team, just to give you an example, a state and local team used to have to go to four or five or multiple different places to get different state matrices, to get the right rates for the different states. A team member came up with a embedded tool within our team’s environment to create a chatbot that basically says, use this as an interface and quickly get all this information in one go. Once we see tools start to get 50, 100, 1,000 users in my state and local example, then it’s our responsibility to put that on the AI internet site. Others get access to it and then start scaling it across the practice overall.

Margaret Burke:

Yeah, and one of the activities that they actually did at the immersion, we had a team use AI to analyze a fictional financial entity as well as their org chart. And they had to identify and prioritize the tax services that we would provide across at least two pieces of our tax practice. And they had to deliver a very clear one-page summary in a four-slide page deck using AI and also speaking to the responsible use of AI.

So again, this is where they’re using the AI tools. They’re learning how to use the AI tools to summarize information, as Krish mentioned, but also the storytelling, right? So really, really important to use that critical thinking, the human skills and the storytelling. We really tried to incorporate that into every hands-on activity that they did during the training.

Krishnan Chandrasekhar:

That’s right, because a big part of it for us, and this is a dance we have to constantly do, is there’s a lot of work that created a lot of kind of friction for our staff. And AI is incredible at taking away some of those frictions. And we’re seeing a lot of that. And that’s what we’ve been talking about.

The other dance we have to do, though, is those frictions actually had learning elements to them. So doing something, even if it was hard, even if it was boring, even if it was repetitive, people learn from that. And so what Margaret and team have to do with us is we need to find those where there was learning and say, how do we replace that? So we’ve taken away one friction, but how do we ensure that we are still teaching them? How are they still thinking about learning? This is where the teaming comes in, ensuring that you’re having the conversations, understanding what reps you’ve taken out of the system so that you ensure that that’s filled in with learning. That’s, I’ll admit, it’s still in the early days. We are still figuring this out, but it’s constantly in the front of our mind. Those reps were important. So if you’re taking away those reps, how are we going to teach?

Jorja Siemons:

Many of these new associates are coming straight out of school. They’re entering a workforce at a time when companies across different industries are investing heavily in AI and also restructuring their businesses. PwC recently said it has shed about 150 jobs across US support roles as it works to reorganize its services, including through using new AI tools.

Margaret, what would you say to early career associates who have received their diploma recently and are wondering how they can best align with skill sets with current AI implementation?

Margaret Burke:

I think, you know, to current students, A, I would say, you know, our goals have not shifted that much from a campus perspective. And B, I would say anything that they can do to upskill themselves, you know, on campus, you know, even outside of on campus. Right. There are several programs and certificate offerings that are available to individuals. I think anything that they can do to upskill themselves certainly helps. But I wouldn’t say that they absolutely, you know, that you absolutely have to have that AI course on your resume, etc., because quite honestly, a lot of the colleges and universities, they’re they’re just getting up to speed as well on what they can offer. And I would say there’s varying degrees of what universities are offering to students right now. So that is where I think if it is available to you and if you can if you can take the coursework, absolutely take it. If you have resources where you can take it externally, I think it certainly helps. But I wouldn’t say at this point that it’s become absolutely necessary.

Krishnan Chandrasekhar:

The other thing I would say is the skills that are never going to go out of fashion is if you’re willing to engage, if you’re willing to opt in and if you’re willing to not stop learning, like don’t think you’ve graduated and so you’re going to stop learning. Those skills are never going to go out of fashion irrespective of where I go. So if you come in with that framework and attitude, there’s plenty to do.

Jorja Siemons:

What has been the largest challenge so far of rolling out and conducting these new trainings?

Margaret Burke:

My visceral reaction to that is the levels of people’s knowledge. So, you know, you’re kind of we brought in three hundred and eighty some odd tax professionals, all with varying degrees of technical ability. So you don’t want it to be boring for some, but you don’t want others to get lost. Right. So you don’t want to start to advance where others get lost. So it’s really keeping the individuals engaged and challenged. So I think that was probably our biggest challenge and one that we’re going to have to work through as we continue to deliver this even to our current employees as well.

Krishnan Chandrasekhar:

Yeah. And from my perspective, it’s doing the balance. But every time there’s new technology, new developments, there’s a tendency and it’s what we want to go heavy on all the cool things we can do with it. From a commercial perspective, the biggest challenge is getting the balance of at the end of the day, what’s the objective of this? We’re still trying to solve client problems. We can’t lose sight of the objectives and the outcomes we have to deliver. So while it’s fun to play with these tools, how do you balance that with you trying to get to an outcome, you’re trying to get an answer. And so are the people still thinking about what is the outcome I have to get and not just what all the cool things I can do with the tool. So it’s again a balance. We want to engage. We want to show them everything you can do, but not lose sight of ultimately clients hire us to get certain outcomes and objectives. So don’t lose sight of that. So that’s the other higher level piece that we have to keep focusing on and reiterating.

Jorja Siemons:

Is it challenging to plan curriculum for these sessions when the technology itself is evolving so fast?

Margaret Burke:

Yeah, it will change. It will absolutely change. So, yes, that is definitely another challenge. I actually think it will change from what we did in October to even what we will roll out in January. There will be pieces of technology that will change. Our own internal tools will change and evolve. So I would say it is a challenge. But I would also say that we have a great team of people that are staying really up to date on what is available, not just externally, but also what we’re building internally for our employees as well as for our clients.

Jorja Siemons:

Let’s say it’s one year from now. What would be a sign that these AI sessions have been successful at training new associates? And what would be a sign that they haven’t been successful at training new associates?

Krishnan Chandrasekhar:

For me, a sign of success, one sign of success is we start to get significant process changes on processes we’ve had for decades, by the way. If in a year from now, I see that important processes that we have had for decades have been enhanced or improved because new staff that came in didn’t have the baggage of the old process and completely rethought a ways because of the training. That would be a sign of success. And I think it’s going to happen. I see it already in the energy in the hallway. So that’s a sign of success for me.

On a side of potential shortfall, if you will, would be what I mentioned a few minutes ago, which is if we get too enamored by the capability of the tool, it has to be a high quality output, perfectly verifiable and solving the client problem. If we lose sight of that and we’ve got all these assets, but the problem solving is still the same or we haven’t improved the value, then we’ve lost something there because of all the investment. That’s my take on it. I know, Margaret, from your lens, I don’t know what you think.

Margaret Burke:

I couldn’t agree more. Part of our reasoning behind this and part of when we went for the investment dollars to do this was to say, listen, we want to inject our culture with this talent, right? And we want to inject them with these individuals that have these skills that can do the reverse mentoring, right? That can from an apprenticeship model can go back to their engagements and teach them what they learned. And that was actually a really important module in each of these immersion sessions in each of the lines of service was how do you now take what you learned over this past week and go back to your engagement teams with individuals that are far more skilled on the tech side than you, but maybe not as skilled as you on the AI side? How do you marry that together and teach the individuals on your teams how to use this technology? And I love what you said, Krish, in terms of they don’t have the baggage of same as last year, right? And this is how we’ve done it for years and years. So, you know, and we talked to them about that. So in terms of hopefully to me, a sign of success is seeing that innovation.

Host (David Schultz):

That was PwC’s Margaret Burke and Krishnan Chandrasekhar talking with Bloomberg tax reporter Jorja Siemons. And that’s it for today’s podcast. You can find up to the minute news on tax and accounting developments at our website news.bloombergtax.com. That website, once again, is news.bloombergtax.com.

Today’s episode was produced by myself, David Schultz, and our editor was Amelia Gruber. From Washington, I’m David Schultz. Thanks for listening.