Rather than solely formal classes, successful companies embed learning opportunities into employees’ day-to-day jobs. This reflects a broader trend to integrate work and learning so tightly that upskilling happens continuously on the job. A Deloitte study notes that rapid changes in work are “making [learning and work] more integrated and connected than ever before,” prompting companies to build “work-centered learning” programs where people upgrade skills in the natural course of their daily tasks. In practice, firms implement regular events and on-demand resources to keep AI learning ongoing. Ally Financial, for instance, holds quarterly “AI Days” where employees hear from experts and see live demos of AI tools, and it has created an internal AI Community with monthly office hours for staff to consult in-house data science experts. This approach encourages employees to bring real work questions and learn AI applications as needs arise. Another example is Adobe, which treats its employees as “customer zero” – staff across departments are enlisted to beta test new AI features (like the Firefly generative AI in Photoshop) and provide feedback. Thousands of Adobe employees have participated in over 30 product betas in a year, simultaneously honing their AI skills and co-creating new tools. By weaving experimentation and micro-learning into routine projects (with cross-functional “AI@Adobe” working groups to support and educate teams), Adobe enables in-the-flow learning. These examples illustrate how constant, applied learning – not just one-time training – helps workers adapt to AI on the fly.

Many leading companies foster circles of learning where employees learn from each other in a collaborative, real-time setting. Accounting firm Crowe created an “AI Guild” open to any employee, after realizing that the spread of AI can feel “scary” or overwhelming for people unfamiliar with the tech. Employees start with a basic generative AI course (covering fundamentals, ethics, and risks), then join the AI Guild to continue learning together through discussions and hands-on practice. Crowe now hosts over 10 such guilds – informal communities focused on various tech and business topics – which serve as “spaces for casual collaboration, networking, and experiential exposure across business units” for any interested employee. This community-first approach supports the way adults learn best: through social engagement, Q&A, and sharing use cases. Similarly, Ally’s AI Community and monthly expert “office hours” enable peer-to-peer learning, and Adobe’s cross-functional AI working group lets teams learn from each other’s experiments. Even on a large scale, companies are leveraging internal social platforms to facilitate knowledge exchange. Rocket Companies (parent of Rocket Mortgage) set up an AI ideas forum called “ChatRKT” where any team member can propose generative AI project ideas and see what others are working on – effectively crowdsourcing innovation and learning. By forming these learning circles, guilds, and forums, organizations tap into employees’ curiosity and create a support network so no one has to learn AI alone.

Companies successful with AI actively involve employees from all departments in exploring AI opportunities, rather than siloing innovation in an R&D lab. At Trek Bicycle, the adoption of AI began with a tech team but quickly expanded to a company-wide task force. Trek’s Advanced Technologies team spent months interviewing every department – from marketing to manufacturing – about how AI could improve their work, gathering employee feedback at all levels. The result was a list of nearly 40 concrete AI use cases addressing real workflow pain points, each developed with input from the relevant department and prioritizing employees’ well-being. This inclusive approach ensured everyone had a chance to contribute ideas and learn. The CIO of Ally Financial echoes this mindset: if AI experimentation is limited to the IT group, it will fail – instead, “the entire enterprise should understand it and be involved in the journey”. In practice, that means forming cross-functional AI teams or “tiger teams” that bring together IT specialists with business-unit staff. It also means creating avenues for bottom-up innovation (like Rocket’s idea portal) so that even front-line employees can suggest AI improvements. By democratizing AI exploration, these companies simultaneously upskill their people – every participant learns about AI applications in context – and surface more diverse use cases than a single team could alone. This strategy builds a culture where employees see themselves as co-creators of the company’s AI future, not passive recipients of automation.

To motivate busy employees to build new skills, some organizations use unconventional, game-based learning and incentives. A standout example is PwC’s “PowerUp” program – a firm-wide AI learning curriculum delivered through a live trivia game. Each month, PwC employees join a friendly quiz competition (“like pub trivia”) featuring questions on AI concepts, firm strategy, and tech trends, all drawn from PwC’s training content. With prizes and bragging rights on the line, the game has drawn over 9,000 participants per month across the U.S. and Mexico. This approach turns AI literacy into a social event and reinforces knowledge through repetition and play. Other firms have hosted hackathons or innovation contests to spur hands-on learning. For instance, Walmart partnered with an external program (Correlation One) to run data science bootcamps for employees – associates could sign up for a cohort-based “Data Science Bootcamp” that prepares them for analytical roles at Walmart. Walmart actively identified employees with aptitude to upskill, enrolling them in this intensive program (free to the employee) which provides a monitored learning experience and even pipelines graduates into new internal AI roles. By making the training experience engaging – whether through gamification, competition, or cohort camaraderie – companies like these increase voluntary participation. The payoff is higher employee engagement and faster skill uptake, as workers are excited to build new capabilities instead of viewing it as a mandatory chore.

Several forward-looking enterprises have invested in comprehensive, multi-year upskilling initiatives to future-proof their talent. Professional services firm PwC is a notable example: back in 2017, PwC launched a $3 billion global “New World, New Skills” program aimed at upskilling its entire workforce in digital and AI skills. Over three years, PwC delivered immersive training in data analytics, automation, and AI to 37,000 employees, significantly raising the organization’s tech fluency. Rather than hire hundreds of external AI experts, PwC chose to retrain its existing staff – a strategy that not only filled skill gaps but also boosted employee loyalty and morale by investing in their growth. A key to PwC’s success was tracking and visibility: they built a culture of learning by actively discussing training goals with employees, highlighting those who completed courses, and monitoring progress so they could encourage wider participation. Over time, PwC even open-sourced parts of its upskilling program to clients and the public, and reports training over 8.3 million total “learner sessions” worldwide (including people outside the firm) in these AI and digital courses. Similarly, Walmart (the world’s largest company) rolled out its “Upskill 2025” effort and other programs to train employees for the future – its Correlation One data science academy is one example showing Walmart’s commitment to moving store and corporate associates into more advanced tech roles. These large programs typically combine online learning platforms, in-person workshops, and certifications. Critically, leadership commitment is high and budgets are substantial, treating workforce upskilling as a strategic imperative. The result is that companies like PwC and Walmart have a talent pipeline ready to leverage AI, rather than being caught in the talent shortage scramble. Indeed, research finds companies that heavily upskill are better able to deploy AI at scale, whereas lack of employee skills becomes a “major bottleneck” for AI adoption in less prepared firms.

Successful AI adopters also recognize that training must address mindset and trust, not just technical skills. Many workers are anxious about AI’s impact on jobs or unsure about its ethical implications. Top companies proactively educate employees on these topics to build trust and buy-in. For example, KPMG’s and Crowe’s AI training modules include dedicated content on AI ethics, responsible use, and risk management. Adobe formed an internal AI ethics committee and review board, including diverse employees, to oversee its AI initiatives and ensure they “enhance, not replace” human creativity. By being transparent about AI plans and emphasizing that employee input is valued in developing AI responsibly, Adobe reports high employee trust – 83% of employees at the “100 Best Companies” (which include Adobe) say people at their organization quickly adapt to change, far above the average elsewhere. Leading companies also communicate clearly how AI will be used. PwC’s leadership, for instance, framed their upskilling as helping employees thrive alongside new tech, not a pathway to layoffs. And at JPMorgan, CEO Jamie Dimon openly wrote to employees that AI’s impact will be as transformative as the steam engine or internet – but assured that while AI may automate some tasks, it will “augment virtually every job” and create new opportunities. This candid messaging, paired with training, helps employees view AI as a tool they will harness (with new skills) rather than a threat. In short, companies winning with AI put just as much effort into culture and confidence-building as they do into technical instruction. This high-trust environment encourages employees to experiment with AI tools and continuously learn, creating a virtuous cycle where the workforce actively accelerates AI adoption rather than resisting it.

the ability to read, analyze, and draw insights from data – is often cited as the single most important skill for the general workforce. AI systems run on data; employees at all levels need to be comfortable working with data outputs to make informed decisions. A global study found that business leaders and employees alike predict data literacy will be the most in-demand skill by 2030, with 85% of executives saying data skills will become as vital as computer skills are today. Data literacy means being able to interpret charts, understand basic statistics or probabilities, and ask the right questions about what the data (or an AI’s analysis of it) is showing. Not everyone will be a data scientist, but everyone will need to use data in their role – indeed, 89% of executives now expect all team members to be able to explain how data informs their decisions. This skill underpins effective use of AI because employees must feed AI systems with quality data and interpret AI-driven insights correctly. As one innovation VP put it, “we need to help [employees] develop skills to add value to the output of intelligent algorithms” – meaning workers should augment AI results with human judgment, context, and critical thinking. Many companies respond to this by rolling out data literacy programs (though the training is often still too concentrated on specialists). Given the gap – only 11% of employees feel fully confident in their data literacy today – boosting data skills across functions (including HR, finance, marketing, etc.) is a priority in future-ready workforce strategies.

For certain roles, deeper expertise in AI development and data science will be needed. The industry is experiencing a surge in demand for AI specialists, including machine learning engineers, data scientists, big data engineers, and related roles. The World Economic Forum projects demand for AI and ML specialists will grow by 40% by 2027, with big data engineers and data scientists increasing ~25%. Companies cannot hire enough of these experts, leading many to reskill internally. For example, some organizations are cross-training software developers or statisticians into AI roles. However, as one tech leader noted, true AI expertise often requires significant specialization and hands-on experience. This means even within technical teams, ongoing training is needed in areas like building machine learning models, understanding AI architecture, and coding for AI. Businesses like IBM have launched internal “AI Academies” and provide courses on topics such as computer vision, natural language processing, and AI engineering. On a broader scale, many Fortune 500 employers estimate ~40% of their workforce will need reskilling in the next 3 years due to AI and automation changes – a figure that includes retraining people into entirely new tech roles. Technical skills in cloud computing, coding (Python, R), data management, and cybersecurity also become more important as AI systems must be implemented and safeguarded – and companies are blending these into their upskilling curricula.

With great power comes great responsibility – AI ethics, governance, and risk management have emerged as must-have knowledge areas. As AI is deployed in sensitive processes (like hiring, lending decisions, or medical diagnosis), employees at all levels should be aware of issues like bias in AI, privacy implications, and compliance requirements. The “future of work with AI” will likely be heavily regulated and scrutinized, so a culture of ethics is a skill in itself. Leading companies already train their workforce on AI ethics: e.g., KPMG’s “Trusted AI” course ensures employees understand guidelines for ethical AI use, and many firms have internal AI ethics committees that involve employees (Adobe’s ethics board is an example). Workers are taught to question AI output, flag potential problems, and make decisions with transparency. This skill set goes hand-in-hand with critical thinking – the ability to not take AI results at face value and to consider broader societal impacts. As generative AI becomes prevalent, employees also need guidance on intellectual property rules, data security (e.g. not inputting confidential info into public AI tools), and company policies for AI use. In short, an employee who is “AI fluent” must also be versed in what not to do with AI. Training programs now commonly include modules on AI ethics and safe usage because employers want a workforce that can help monitor AI’s proper use and maintain trust with customers and regulators.

Ironically, the more work we automate, the more human skills gain importance. Abilities that AI cannot easily replicate – creativity, innovation, problem-solving, communication, leadership, and empathy – are increasingly crucial. These are the skills that allow employees to partner effectively with AI. For instance, AI can churn through data and generate insights, but a human with domain expertise and creativity will envision how to use those insights in new products or strategies. Critical thinking and curiosity drive employees to ask the right questions that direct an AI’s analysis. Collaboration and emotional intelligence remain key in workplaces where teams include both humans and AI tools; employees must excel at the things like teamwork, customer understanding, and ethical judgment that technology alone can’t handle. Many future-of-work reports highlight adaptability and lifelong learning as essential skills – since job roles will evolve continuously, workers need the resilience and mindset to constantly learn new things. Companies such as Unilever and IBM have explicitly started to evaluate and train for “learnability” and soft skills, anticipating that technical skills may change but these human capabilities will endure. Even in tech-heavy fields like banking, leadership is training managers on how to integrate AI into decision-making while still motivating and mentoring their teams. In summary, the consensus is that hybrid skills will define the successful AI-era employee: a blend of technical/digital proficiency and strong soft skills. As one World Economic Forum insight put it, employees will be expected to augment machine intelligence with human creativity and critical thinking – so both aspects must be developed.