Most companies are teaching prompts and platforms. PwC is rethinking development itself.
The corporate world has no shortage of AI courses, boot camps, and crash programs. Employees are being taught how to prompt, automate tasks, and navigate new tools.
The harder challenge, however, may be helping people combine those new capabilities with the judgment, communication, and curiosity that technology cannot replace. In other words, using AI is one skill. Using it well is another.
PwC is trying to address that challenge through its Learning Collective, a workforce initiative designed to blend AI capability with human skills in day-to-day work.
Read more: PwC chief urges staff to embrace AI to stay relevant
Margaret Burke, Talent Acquisition and Development Leader at PwC, said the company identified 15 human skills and 15 AI technical skills, then built its training model around developing them together.
“We wanted to make sure that whenever a technical skill is being taught anywhere within the firm, a human skill has to be taught alongside it,” Burke said. “When we were teaching prompting, we wanted to make sure that we were really amping up critical thinking so that whatever came out of that prompting, you would have a critical eye to the output.”
At the center of that strategy is the idea that AI capability should be paired with human skills such as critical thinking, curiosity, and storytelling so employees can use the tools more effectively, Burke said. She prefers the term “human skills” to “soft skills,” arguing the older label carries a negative connotation. She added that those capabilities could become a key differentiator for workers.
However, teaching those skills can be more complicated than teaching software, Burke acknowledged.
“Some of those human skills people think are innate, either you have them, or you don’t,” she said. “But sometimes it’s just teaching them the level of awareness of when they need to lean into those human skills a little bit more.”
The workday becomes the classroom
PwC’s learning model is changing as well. Rather than relying on scheduled classroom sessions, the firm is moving development closer to the work itself through hands-on challenges, real-world scenarios, and AI-enabled support that appears in the daily flow of work, according to Burke.
“Now you will almost never see a training class where there are not hands on keyboards, where PCs aren't up in front of them or some type of tools,” she said. “We’ve done vibe coding sessions, prompting challenges, scaffolding challenges. It’s all literally hands on the keyboard so that we’re working through problems and challenges together.”
PwC is also piloting AI coaches and assistants inside platforms such as Microsoft Teams and Copilot. Employees can rehearse client meetings, test ideas, and receive feedback before the real conversation begins, Burke said.
“We’re deploying agents that can help have client conversations for our directors or our partners,” she said. “They can literally practice the client conversation with a custom AI agent before they go to the client.”
Instead of pulling employees out of work to learn, PwC wants learning to happen while the work is happening.
AI can give the answer. People give it meaning.
These changes may be felt most sharply by early-career employees. PwC began its recent AI learning push with entry-level hires because those roles were likely to feel the effects first, particularly where repetitive work is concerned, Burke said.
“We really started in earnest about a year ago with our entry-level hires because that was kind of low-hanging fruit, and we knew that AI was going to change their jobs probably most readily, probably most immediately,” she said.
Read more: Shadow AI rises as employees outpace workplace controls
AI is likely to absorb more rote and mechanical tasks, which could place greater value on judgment, communication, and client-facing skills earlier in employees’ careers, according to Burke.
That is one reason PwC is placing heavier emphasis on storytelling. In a world flooded with AI-generated analysis, the ability to turn information into a clear narrative may become more valuable.
“While AI is going to help with the output, how we go back and tell the story of what’s happening and how it will impact the business to our clients changes immensely,” Burke said. “Our ability to story-tell becomes much more important.”
The hardest part? AI keeps changing.
Rolling out new learning models is one challenge. Keeping pace with the technology itself is another.
The rapid speed of change has been one of the toughest parts of the effort, Burke said, with tools evolving quickly and priorities shifting just as teams get comfortable.
“We’ll get up to speed on one tool or one area that we want to focus on, and then it moves so fast. You have to constantly be on the cutting edge,” she said.
The speed of AI change is also forcing companies to rethink slower, traditional training cycles, she said, because by the time a course is designed, approved, and delivered, the technology may already have changed.
“We’re not really putting people into a traditional classroom setting anymore. It’s much more about how we learn while doing, how we learn from one another, and how we keep pace as things evolve,” Burke said.
To stay current, PwC has a dedicated innovations team focused on learning new technology. They’re also actively piloting multiple tools across coaching, workflow learning, simulations, and content creation, she said.
That pace is also affecting employees.
“I think everyone has a little bit of anxiety around AI,” Burke said. “Everyone thinks they’re behind, and I think it’s because it’s constantly changing.”
Lessons for employers still figuring it out
Asked what other organizations should keep in mind as they build their own AI learning strategies, Burke urged leaders not to wait for a perfect plan.
“Don’t be afraid to make a mistake. Don’t be afraid to lean into the tools,” she said. “While it can seem intimidating at first, it’s a great place to learn.”
Curiosity also matters, especially for leaders trying to guide others through rapid change, she added.
“I think it does take a little bit of that curiosity,” Burke said. “We get on once a week and talk about what we learned this week, because if someone shows me how to do something, that’s so much better than me reading it.”
Her broader message was that adoption is not only about tools. It is also about confidence, peer support, and helping employees see where they fit in an AI-enabled future.
“Right now, we see it as a career enhancer and a career elevator, not a career eliminator,” Burke said.
