AI & the Workforce: Strategic Implications
credit: wenich-mit
This article addresses the Human Factors: Fear, Trust, Workforce Impact hurdle in my Moving Forward with AI primer article.
AI: Not a Tech, but A Leadership Challenge
As organizations face growing pressure to adopt AI, recent Stanford and BCG surveys show 78% now report active AI usage, up sharply from 55% just last year. At the same time, nearly half of employees in AI-intensive organizations (46%) express concern about job security than those in less AI-focused firms. The anxiety is real. Many executives tell me that preparing people for change — shifting mindset and culture — is the biggest hurdle.
How can leaders navigate this revolutionary shift that is unfolding faster than most organizations are structurally or culturally ready for?
While the right path forward should depend on each organization’s unique circumstances, a few key concepts and frameworks are especially useful based on insights from strategy, human capital, and organizational change leaders as well as my own research. We will explore:
A. How AI Is Reshaping Work and The Workforce
B. From Resistance to Readiness: Starting from The Personal
C. Rethinking Workforce Strategy
D. Leading in The Age of AI
E. The Path Forward
A. How AI Is Reshaping Work and the Workforce
The fear of AI replacing humans remained remote until recently. Today, transformer-powered generative AI such as ChatGPT increasingly takes on cognitive and creative tasks, and Model Context Protocol (MCP) allows AI models to connect more easily with external tools and data sources. As AI evolves beyond assistant tools into more autonomous “agentic AI” systems, organizations find themselves having to rethink workflows once reserved for knowledge workers. These advances bring new opportunities for automation and augmentation while raising urgent questions about workforce strategy.
Yet, many leaders lack the time or support to fully grasp AI’s realities. They over-rely on optimistic promises of consultants or solution providers. This reality gap often explains why AI initiatives underdeliver, and is especially harmful when executives jump into headcount reduction before rethinking workflows.
1. How AI Works Differently Reshaping Work
Unlike traditional rule-based systems, modern AI models recognize patterns, weigh probabilities, and make inferences. This lets machines do what traditional computing cannot, for example, predicting trends or personalizing customer experiences.
Research shows some models already outperform humans in reading comprehension and image recognition in narrow benchmarks. Handwriting and speech recognition have reached human level, and predictive “reasoning” – often seen as a final frontier – has made substantial leaps in narrow numerical and logical reasoning. With these advances, AI is no longer just a niche tool for data scientists and engineers. It’s becoming a broadly accessible force reshaping knowledge work. This presents significant opportunities that come with difficult leadership challenges: How do we realign roles and support the workforce through this shift?
2. The Moravec Paradox
“What seems easy is actually hard, and what seems hard is actually easy,” MIT Professor Hans Moravec famously pointed out, referring to how tasks that are intuitive for humans are surprisingly difficult for machines to replicate, and vice versa. While AI excels at detecting patterns and processing vast amounts of data at greater speed than humans, it struggles when the context is complex and nuanced judgment is required. In other words, “Meaning-Making” to determine “the subjective value of a thing” remains human, as Professor Vaughn Tan explained in The Meaningmaking Lens on AI.
Consider software development, once the domain of specialists. Large Language Models (LLMs) now allow non-tech users to “prompt” in plain English, changing who can build automated solutions, how quickly, and at potentially much lower costs.
So when we think about adopting AI to optimize business value, it’s not a matter of humans vs machines – it’s about harnessing respective strengths to deliver outcomes neither could achieve alone. Several studies have shown that human + AI hybrid teams outperform humans or AI working alone in fraud detection, medical diagnosis, and cybersecurity prevention.
3. Automatable and Augmentable Task Framework
How do leaders go about reshaping their workforce to achieve the best balance of human + AI operating models? The framework of Automatable and Augmentable Tasks, popularized by the World Economic Forum’s analysis of the impact of LLMs, may be useful. The report suggests routine and uncreative tasks are automatable. Tasks requiring contextual understanding, critical thinking, or relationship-building are augmentable.
For example, commercial insurance underwriting can be augmented by AI that synthesizes data from multiple sources into an initial risk profile. This enables underwriters to handle more policies (productivity gains) while applying deeper judgment to price more complex policies (new possibilities).
As technology advances, tasks that are augmentable today may gradually shift toward full automation. Leaders need to anticipate this “task creep” proactively, creating incentives for employees to adapt, experiment, and focus on uniquely human contributions that machines cannot easily replicate. The strategic and emotional risks of upskilling roles that may soon be automated cannot be overstated. Aligning skill development with both today’s immediate needs and tomorrow’s evolving roles demands adaptability from both managers and employees.
4. Human-In-The-Loop (HITL)
As leaders think about redesigning AI-powered workflows involving humans and machines, the Human-In-The-Loop (HITL) concept is critical for decision-making tasks with regulatory, business, or risk implications.
Human-Led with AI Assistance: AI analyzes data and generates recommendations for humans to decide, such as Zurich’s claims analysis AI.
Shared Autonomy: AI handles routine tasks automatically with humans supervising and stepping in for exceptions, such as Lemonade’s AI bots for policy admin processing.
AI-Led with Human Supervision: AI acts independently within set parameters, with humans monitoring and intervening as needed, as in low-risk claims like a stolen bike.
Full Automation: AI handles tasks end-to-end, as in throughput travel claims.
Given the impact on individual consumers and the known limitations of LLMs, such as hallucinations, when and how to keep a human in the loop is a critical design decision.
Since most organizations are built around traditional computing, introducing AI inevitably raises the question: How do we reimagine work and evolve roles to deliver smarter, better outcomes?
B. From Resistance to Readiness: Starting from the Personal
Ironically, in the age of AI, humans are being called on to be creative, ask good questions, and empathize. Yet in many organizations still run like machines, employees are discouraged from developing these capabilities. The issue is therefore not simply about humans versus machines; rather, it’s about how we define and expect work to be done.
In preparing for an AI-powered future, leaders should first truly understand the strengths and limitations of AI and think about what it means for employees at all levels. In genuinely embracing the power of combining humans’ and machines’ respective strengths, leaders can then create space and psychological safety for employees to experiment and grow, replacing fear and anxiety with curiosity.
Resistance to change can also come from a lack of clarity and trust. When people feel a sense of purpose, they’re more motivated to perform at a higher level and more open to rethinking work, especially when they see that their contributions are valued and recognized. For example, when employees see that AI tools help them understand customers better, they can go the extra mile and provide better service while getting satisfaction from helping customers personally.
Employees engage more when they see that embracing new ways of working leads to career progression. For example, insurance underwriters who adopt AI tools proactively could be rewarded with more complex portfolios or leadership pathways.
Leaders who address fear and resistance at the human level, provide purpose, and build trust by encouraging curiosity can turn resistance into stronger employee engagement, which is the foundation for sustainable AI adoption.
“In many cases, the cultural shift required is not about trusting AI, but about trusting humans.” – Chris Paton,
C. Rethinking Workforce Strategy in the Age of AI
Once leaders embrace the Human + Machine and the Automatable vs. Augmentable Task framework, they can start to rethink ways lead the shift in their organizations.
1. Workforce Strategy Must be Central to AI Adoption
People by nature resist change, especially when it feels threatening. Recent surveys have revealed that while executives are bullish on AI, many employees remain skeptical or even resistant, with some going as far as sabotaging tools they find unhelpful or threatening to their job or identity. Good leaders understand this. They also recognize that employees often see where efficiency gains and new opportunities exist.
Moreover, many organizational structures have been built around siloed control, with systems unable to respond to the fluid and dynamic nature of AI-enabled work. To unlock AI’s potential, leaders inevitably will need to engage employees across all levels and reimagine a workforce involving both humans and AI in its many shapes or forms.
Given the powerful potential of human + AI collaboration, leaders should focus on elevating employees as value creators rather than costs to be cut. The mixed results of offshoring and robotic processing automation show that rushing to reengineer with fewer workers rarely delivers the promised results.
In markets facing a shrinking workforce or talent shortages, human-AI collaboration is particularly impactful, helping to maintain service levels and even expand capacity, allowing smaller teams to achieve more.
Beyond productivity gains, AI can take on routine analysis so decisions can move closer to the front line — speeding tempo, testing ideas faster, and freeing humans to spend more time with colleagues and customers. This strengthens collaboration and allows employees to deliver quicker and more responsive service.
2. Address the “Middle Squeeze”
Middle managers play a pivotal role in AI adoption. They are often tasked to decide what to deploy AI for and how to manage new risks and responsibilities. These decisions aren’t just technical - they involve understanding risks, employees’ capabilities, and impacts on customers. Yet they’re facing growing pressure from above to implement change while having to manage teams anxious about their future. This “middle squeeze” is real. A recent BCG report found nearly 60% of middle managers feel unprepared to lead AI initiatives.
Without clear support and guidance, middle managers often become a bottleneck. Empowering them with the right tools, clear direction, and real authority, and supporting their shift from “command-and-control” to “coaching” mindset is a critical leadership investment for sustainable AI adoption.
But capability alone isn’t enough if legacy hierarchies and outdated performance measures constrain managers from experimenting. To fully enable the middle, KPIs, incentives, and reporting lines must be realigned to reward AI-enabled improvements.
3. Redefine ROI: Measuring the Value of Human-Centric AI Adoption
Traditional ROI measures, particularly those focused narrowly on labor costs, miss the deeper, longer-term value AI can deliver.
AI models often require multiple iterations, and realigning roles and responsibilities is rarely fast.
While some cases, such as chatbots, may deliver quick wins, more impactful applications involving predictive analytics or autonomous decision-making often take longer to yield returns, particularly in insurance with long policy cycles.
AI brings intangible benefits that are difficult to capture with traditional metrics — how do you quantify faster decision-making, deeper insights, and the strategic advantage of a more adaptive and innovative culture?
To evaluate AI’s true impact, leaders should broaden the traditional ROI lens to consider the interconnected play of gains in quality, decision speed, customer loyalty, and talent engagement. For example, when employees feel valued and supported through change, they are more likely to deliver high-quality service that strengthens customer relationships and drives revenue.
Ultimately, understanding how AI reshapes work and creates new value is key to developing metrics that measure what truly matters.
4. AI Literacy for Workforce Readiness
AI literacy is fast becoming a non-negotiable skill. Some hiring managers now see it as even more valuable than prior job experience. Many organizations are investing in AI training and upskilling, but much of it remains tactical, focusing on tools.
AI readiness needs to move beyond training checklists to create ongoing opportunities for employees to learn by doing. This means helping them truly grasp AI’s capabilities and limitations and build confidence in interpreting outputs and questioning underlying data and assumptions. For example, how do AI models assess risk and flag anomalies?
Leaders play a crucial role in setting the direction for AI literacy: they need to clearly explain how AI fits into the organization’s overarching strategy, how data strategy will align, and crucially, why it matters.
An AI-literate workforce means all employees share baseline understanding of how AI works, a common vocabulary across tech and non-tech, and the confidence to engage and contribute. It also requires leaders to show up and set the tone for a culture of continuous learning.
5. Cross-Generational Knowledge Sharing
As roles evolve, so must hiring. Today’s young workers enter the job market fluent in digital tools with immense potential to drive innovation. Organizations that harness the ingenuity of the younger workforce and foster cross-generational knowledge-sharing will be better positioned to stay relevant and competitive.
JPMorgan Chase understands this. They built AI tools for initial analysis, and trained junior analysts to ask better questions, test assumptions, and learn decision context from seasoned analysts. They recognize that well-designed machines can handle much of the routine tedious work faster and more reliably than junior analysts, but AI’s output still needs validation, and final decisions still require human judgment.
6. AI-Enabled Talent Development
While AI is reshaping the workforce, it has also made personalized upskilling possible. Johnson & Johnson, for example, uses AI to quantify skills and proficiency and identify gaps to enable targeted training. They also use a structured approach to define future-ready skills (using aggregated employee data to ensure privacy) and heat-maps for workforce planning.
AI-enabled upskilling can adapt to individual employees’ needs and motivations. Generative AI tools, for example, have been found to raise the performance of lower-skilled and less experienced workers and improve productivity for average performers, but don’t always boost top-tier talent.
D. Leading in the Age of AI
1. Addressing Expectation Gaps
AI’s complexities have created significant gaps in expectations and knowledge between leaders and employees, according to a survey of 800 knowledge workers and 800 executives. How should executives and boards lead in the age of AI?
For organizational change to take hold, Edgar Schein’s framework on primary and secondary mechanisms is useful. Systems, processes, and training are Secondary Mechanisms. Primary Mechanism — what leaders pay attention to, invest in, and reward — sends the strongest signals. In AI adoption, if leaders say AI is a priority but fail to fund training or reward experimentation, the culture won't shift, and the change won’t last. They must clearly define what the organization seeks to achieve with AI, how different technologies work together, and how every level is expected to contribute to creating value with AI.
2. Building Trust
AI tends to amplify what already exists – for better or worse. Given AI’s uncertainty and complexity, transparency matters. leadership and the board should openly talk about the organization’s AI strategy, the roles AI will play, and how human contributions remain critical. It also means discussing hard truths: some roles will evolve, others will disappear. Avoiding hard conversation erodes trust. This is, of course, easier to do in high-trust cultures. If that trust isn’t there yet, leadership will need to invest in trust-building along with technology adoption.
E. THE PATH FORWARD
As with past high-impact innovations such as electricity and the internet, how AI reshapes work will depend less on technology and more on the vision and choices of leaders. My hope is that the concepts discussed in this article will help leaders approach AI adoption with clarity and purpose and empower employees with better tools and more timely and useful information to pursue new growth and shared prosperity. Specifically,
Make Workforce Strategy Central to AI Adoption. Redesign work to enable humans to be their best. Regularly reassess what tasks remain augmentable and what are becoming automatable while empowering employees.
Transform Resistance to Readiness with Curiosity and Purpose. Invest in AI and data literacy. Create psychological safety and encourage employees to reimagine how they contribute and incentivize them accordingly. Motivated employees bring energy to customers and drive results.
Support Middle Managers Leading the Shift. Middle managers are the ones who turn vision into reality. Equip them with guidance, tools, and authority to turn strategy into real-world outcomes.
Leverage Cross-generational Strengths. Pair seasoned employees with industry and institutional knowledge (e.g., veteran underwriters guiding risk judgment) and younger employees fluent in AI and digital tools (e.g., junior underwriters managing data and flagging patterns) to unlock innovation across generations.
Invest in Trust as a Competitive Advantage. Be transparent and explain constantly how AI is used, where human judgment still matters, and what principles guide design choices. Earning trust with employees, customers, and regulators creates a durable competitive advantage in a fast-evolving world.
Rethink How Success is Measured. Expand ROI beyond traditional metrics and short-term efficiency gains. Value innovation and the long-term edge of an AI-ready workforce able to move faster, decide better, and delight customers at lower cost.
The path forward calls for courageous leadership, empowered humans, and reimagined work. When done right, this shift to human-centered AI adoption can lead to smarter ways of working, more resilient organizations, and a purposeful future worth striving for.
Ready to move forward or have questions about purposeful AI adoption? Get in touch.
* Ichun Lai founded Propel Global Advisory LLC focusing on accelerating the thoughtful and responsible adoption of AI technology in financial services