With decades of experience helping organizations navigate major technological shifts, HRTech expert Ling-Yi Tsai has become a leading voice in the integration of artificial intelligence into the workplace. Specializing in HR analytics and the strategic implementation of technology across the entire employee lifecycle, she argues that HR’s role is no longer about managing programs but about architecting the very future of work. In our conversation, we explored how HR leaders can move beyond simply adopting AI tools to fundamentally redesigning work itself. We discussed the critical need to build new frameworks for performance, establish systems that ensure trust and accountability, and intentionally design AI integration to close, rather than widen, the opportunity gap for employees. Ultimately, our discussion centered on how to transform AI from a source of anxiety into a catalyst for human capability and growth.
Research shows many organizations struggle to scale AI beyond pilot projects, a gap often rooted in work design. How can HR leaders move from merely adopting AI tools to fundamentally redesigning roles and workflows to close this execution gap and ensure enterprise-wide success?
It’s a gap I see constantly, and it’s deeply frustrating for leadership because they invest in powerful technology only to see it stall out. The McKinsey data is spot-on: only about 20 percent of companies manage to scale AI effectively. The core issue is that they treat AI as a shiny new object to be plugged into existing, often outdated, processes. That approach creates immediate friction. You can’t just hand a team an AI-powered analytics tool and expect magic. You have to break down the work itself into dynamic tasks and ask, “What is the machine best at, and where is human judgment indispensable?” This means moving from static job descriptions to a fluid model where people and AI collaborate. When HR leads this redesign, the transformation is palpable; when they don’t, you just get confusion and a failed pilot project.
As AI handles more routine tasks, traditional metrics like output can become misleading. What new frameworks should HR develop to measure performance, and how can they effectively evaluate uniquely human skills like judgment and collaboration in an AI-supported environment? Please provide a specific example.
This is one of the most critical shifts HR must navigate. Measuring purely on output or efficiency becomes almost meaningless when an AI co-pilot is doubling someone’s speed on routine tasks. The focus has to pivot to the quality of human-led decisions with AI support. We need to measure judgment and collaboration. For instance, consider a supply chain planner. Previously, they might have been measured on the number of orders they processed. Now, with AI forecasting demand, their real value is in managing exceptions. Did they correctly interpret the AI’s warning about a potential disruption? How effectively did they collaborate with logistics and sales to create a contingency plan? Performance is no longer about the volume of their work but the quality of their intervention. HR’s role is to build a framework that rewards that sophisticated judgment, not just the button-clicking.
Building trust in AI is critical. What practical systems, like transparent escalation paths, should HR build to ensure accountability when AI influences hiring or promotions? Describe a process for employees to safely question an AI-driven decision and for leaders to effectively intervene and correct errors.
Trust can’t just be a talking point; it must be engineered into your systems. It feels solid and reliable. Imagine an employee is flagged by an AI-driven performance management tool for a potential decline in productivity. The first step in a trustworthy system is transparency: the employee is immediately notified not just of the flag, but of the key data points that triggered it. The second step is a simple, clearly communicated escalation path. The employee should have access to a one-click button in their HR portal that says, “Request a human review.” This action should trigger a formal review by their direct manager and an HR business partner, with no penalty or prejudice. This process shouldn’t feel confrontational. It’s a system check. The final, crucial step is the feedback loop; if an error is found, the data is used to retrain and improve the algorithm. This demonstrates that human oversight is real and that the organization is committed to fairness, making the technology a tool, not a verdict.
There’s a growing “meaning gap,” where managers feel more optimistic about their careers than frontline workers. As AI reshapes work, how can HR intentionally design opportunities and bias-monitoring systems to ensure AI closes this gap rather than widens it, promoting fairness across all levels?
That “meaning gap” highlighted in the PwC survey is a real danger. If we’re not careful, AI could become a massive engine for inequality, automating away entry-level opportunities while creating hyper-specialized roles only accessible to a few. HR must be the architect of fairness here. This starts with building active bias-monitoring directly into AI-enabled systems for hiring and promotion. It’s not enough to deploy the tool and hope for the best. We need dashboards that show us, in real-time, if the AI is favoring candidates from certain backgrounds or overlooking talent in frontline roles. More importantly, HR can use AI to expand access. We can design systems that identify high-potential frontline workers based on their skills and capabilities—not their job titles—and proactively suggest development pathways and mentorship opportunities. This intentional design work turns AI from a potential barrier into a bridge, ensuring that the opportunities it creates are distributed equitably across the entire organization.
To avoid anxiety, employees need to understand how AI supports them. What specific design principles should HR follow to create psychological safety? Can you share a step-by-step approach to implementing AI in a way that builds employee confidence and a sense of “superagency”?
Psychological safety doesn’t come from posters on the wall; it’s a direct result of good, clear design. First, you must always prioritize clarity over cleverness. Every employee needs to understand exactly where AI is being used and where human judgment remains the final authority. This means no “black box” decisions. Second, frame the implementation around augmentation, not replacement. The narrative should always be, “This tool will handle the repetitive tasks so you can focus on the strategic work you were hired to do.” Third, co-create with your employees. Before a full rollout, run workshops with the teams who will use the technology. Let them help design the workflows. This builds ownership and demystifies the process. This approach cultivates what McKinsey calls “superagency,” where employees feel empowered by the technology, not threatened by it. They feel like they have superpowers, able to achieve more than they ever could alone, because the system was designed to enhance their capabilities from the very beginning.
The idea of learning is shifting from static training programs to dynamic capabilities embedded in daily work. What does this look like in practice? How can HR create systems where skill development happens continuously, in context, and helps employees see clear pathways for internal mobility?
The old model of sending employees to a two-day training course is broken. In the AI era, learning must be like breathing—continuous, contextual, and integrated into the flow of work. In practice, this means using technology to identify skill gaps and opportunities in real time. For example, an AI-powered system might notice that a marketing specialist is increasingly working on projects that require data analysis. Instead of waiting for a formal review, the system could immediately serve up bite-sized learning modules on data visualization or suggest a short-term gig on the analytics team. This makes skills visible across the entire organization. Suddenly, an employee’s capabilities are no longer locked within their job description. They become a dynamic profile of what they can do, and the system actively helps them see clear, tangible pathways to grow and move within the company. This transforms learning from a scheduled event into a constant state of becoming.
What is your forecast for the future of work?
The future of work will not be defined by the sophistication of our AI, but by the wisdom with which we design work around it. My forecast is that by 2026, the most successful and resilient organizations will be those where HR has fully stepped into the role of architect. They won’t be managing HR programs; they’ll be stewarding the very operating system of their company. We’ll see a clear divide between organizations where AI has eroded trust and widened the “meaning gap,” and those where it has genuinely expanded human capability and fairness. Progress won’t be measured by how many tasks we’ve automated, but by what our people are able to achieve, together, with technology as a true partner. That outcome isn’t inevitable—it will be the direct result of intentional, human-centered leadership today.