10/19/2026
The AI Revolution Has a Leadership Problem
Why billions in technology investment are producing extraordinary productivity gains—but not necessarily extraordinary business results.
“The true measure of leadership is not how quickly we adopt new technology, but how effectively we turn its potential into lasting value.” — Scott Podvin
Artificial intelligence is transforming what people can accomplish. Yet one of the most consequential questions of the AI revolution remains surprisingly difficult to answer:
If employees can accomplish more in less time, why aren't businesses consistently reporting corresponding improvements in profitability?
The distinction matters. A company can purchase powerful technology, automate thousands of tasks, and celebrate impressive productivity statistics without creating a sustainable competitive advantage.
Technology can expand an organization's capabilities. It cannot, by itself, determine whether those capabilities produce economic value.
That responsibility belongs to leadership.
The Great AI Contradiction
The adoption of artificial intelligence has advanced far faster than many organizations' ability to measure its financial contribution.
McKinsey's 2025 State of AI survey found that approximately 88% of respondents reported AI use in at least one business function. Yet only about 6% met the firm's definition of AI high performers—organizations reporting significant value from AI and attributing at least 5% of enterprise EBIT to its use.
These figures do not establish that the remaining organizations have failed. Many investments are still developing, and some benefits take years to materialize.
They do, however, expose a significant distinction between using technology and capturing its economic value. [1]
A business can improve individual productivity while simultaneously absorbing higher technology expenses, integration costs, employee training requirements, and quality-control obligations.
The result can be an organization that operates faster without generating proportionately greater earnings.
For investors and corporate executives, the relevant question is no longer simply whether AI works.
It is whether the organization can convert what works into durable financial performance.
When Productivity Becomes Real
Research demonstrates that artificial intelligence can create substantial productivity improvements under the right conditions.
In Generative AI at Work, Erik Brynjolfsson, Danielle Li, and Lindsey Raymond examined the introduction of an AI assistant among 5,179 customer-support agents.
Their research found that access to AI increased productivity, measured by issues resolved per hour, by approximately 14% on average. Less-experienced and lower-skilled employees experienced improvements of approximately 34%.
The researchers also found evidence of improved customer sentiment, greater employee retention, and accelerated learning. [2]
These findings are important for reasons that extend beyond efficiency.
AI can help transfer institutional knowledge from experienced employees to those who have not yet developed the same expertise. It can reduce the time required to become productive and potentially broaden access to higher-value work.
That is a genuine empowerment opportunity.
But the study measured operational productivity, not a corresponding 14% improvement in company earnings.
The distinction is fundamental.
If a customer-service department resolves more inquiries per hour, the financial benefit depends on what happens next. Does the organization reduce overtime? Improve retention? Serve additional customers without proportional staffing increases? Generate more revenue? Or simply absorb the productivity improvement without changing its operating structure?
Productivity creates potential value. Management determines how much of that value is realized.
Klarna: A Case Study in Economic Ex*****on
Financial technology company Klarna offers a particularly instructive example.
In February 2024, Klarna announced that its AI-powered customer-service assistant had handled 2.3 million conversations during its first month, representing approximately two-thirds of its customer-service chats.
The company reported that average resolution times declined from approximately 11 minutes to less than two minutes, while repeat inquiries fell by 25%. [3]
Those were impressive operating results. But the more consequential question was whether they translated into measurable financial benefits.
In subsequent securities disclosures, Klarna reported approximately $39 million in cost savings attributable to its AI assistant during 2024. The company also estimated that the system performed work equivalent to more than 700 full-time agents. [4]
That is stronger evidence than a productivity demonstration alone.
Nevertheless, even this example requires careful interpretation. The savings are company-reported estimates, not an independently isolated calculation of AI's contribution to total enterprise earnings.
Klarna's broader financial performance reflects numerous factors, including business growth, credit performance, operating expenses, and market conditions.
The lesson is not that every company should replicate Klarna's approach.
It is that leadership must establish the connection between a technological intervention, measurable operational improvements, and financial outcomes.
Customer experience must also remain part of that calculation. A system that reduces service costs but damages customer relationships may create an apparent short-term gain while undermining long-term enterprise value.
The most successful implementations must balance efficiency, service quality, and sustainable economics.
The Missing Link Is Organizational Design
Why do some organizations translate AI adoption into meaningful economic performance while others struggle?
McKinsey's research identifies a revealing pattern.
In its 2025 survey, fundamental workflow redesign was the organizational attribute most strongly associated with enterprise EBIT impact from generative AI. Yet only 21% of respondents whose organizations used generative AI reported fundamentally redesigning at least some workflows. [5]
This suggests that the principal obstacle is not always technological sophistication.
It may be organizational inertia.
Consider a company where customer inquiries pass through several departments, information is entered repeatedly, approvals are delayed, and employees spend significant time correcting incomplete records.
Introducing AI into individual tasks may accelerate portions of that process.
But if the underlying workflow remains fragmented, the company may simply move information more quickly between inefficient systems.
The alternative is to redesign the entire process.
Which activities should be automated? Which decisions require human judgment? Where should information be captured? Who is accountable for errors? How will the organization measure improvements in cost, quality, and customer satisfaction?
These are leadership decisions.
They require executives to challenge established operating procedures, reallocate resources, redefine responsibilities, and sometimes confront organizational resistance.
Technology alone cannot resolve those questions.
The organizations that create lasting value will not necessarily be those that automate the most activities. They will be those that eliminate the greatest amount of unnecessary work while improving the outcomes that matter.
The Financial Test Every AI Investment Should Pass
The economics of artificial intelligence deserve the same scrutiny applied to any significant capital investment.
Consider a hypothetical company evaluating an AI initiative expected to generate $500,000 in annual operating savings.
Assume annual technology, integration, oversight, and support costs total $250,000.
At first glance, the initiative appears attractive.
But what happens if the organization realizes only 70% of the projected savings?
The arithmetic changes considerably:
Annual financial measure Amount
Projected gross savings $500,000
Actual realization 70%
Realized savings $350,000
Annual operating and implementation costs ($250,000)
Annual net benefit $100,000
Simplified annual return on costs 40%
The initiative still generates a positive annual net benefit under these assumptions. But its realized contribution is substantially lower than the original projection.
At a 50% realization rate, the annual net benefit falls to zero.
This simplified illustration excludes upfront capital expenditure, financing costs, taxes, and the timing of cash flows. A complete investment analysis would incorporate those factors through discounted cash flow and sensitivity analysis.
It nevertheless reveals the leadership challenge.
A forecast is not a result. A demonstration is not an operating system. And a productivity improvement is not automatically an economic return.
Organizations should establish baseline performance before implementation, measure actual results afterward, and regularly reassess whether the initiative continues to justify its cost.
For investors, the discipline is familiar: capital must be evaluated against risk-adjusted economic returns, not the excitement surrounding its deployment.
The Human Capital Paradox
The AI revolution also raises a question that financial statements alone cannot answer.
Who benefits when employees become more productive?
An organization might use AI to reduce staffing expenses. Alternatively, it might equip existing employees to manage more complex assignments, improve customer relationships, or develop new products and services.
Both approaches can create financial benefits, but their long-term implications may differ.
A company that treats AI exclusively as a labor-reduction mechanism risks overlooking the value of institutional knowledge, employee development, and customer trust.
A company that treats AI as an opportunity to expand human capabilities may discover new sources of revenue and competitive advantage.
Neither outcome is guaranteed.
The customer-support research suggests that AI can disproportionately benefit less-experienced employees by helping them acquire the practical knowledge of more accomplished colleagues.
That creates an opportunity to reduce barriers to advancement.
But realizing that opportunity requires training, thoughtful job design, and meaningful pathways for employees to assume greater responsibility.
The central leadership question is not simply how many hours technology can eliminate.
It is what people and organizations can accomplish with the capacity those hours create.
Who Captures the Economic Dividend?
Even when AI produces genuine productivity gains, the benefits do not necessarily flow to a single group.
Shareholders may receive higher profits. Customers may benefit from lower prices or better service. Employees may receive higher compensation or gain access to more valuable work. Technology providers may capture additional revenue through software, infrastructure, and services.
The distribution depends on competition, market power, organizational decisions, and the availability of complementary skills and investments.
This is why productivity growth and broadly shared prosperity are related but distinct outcomes.
A company can become more efficient without necessarily becoming more equitable. Conversely, organizations that invest in workforce capabilities may strengthen both performance and opportunity.
Leadership requires acknowledging these trade-offs rather than assuming that technological progress will distribute its benefits automatically.
The challenge is to develop economic models in which innovation strengthens the enterprise while creating sustainable value for the people who contribute to and depend upon it.
Five Questions Every Leader Should Answer
Before approving another major AI initiative, boards, executives, and investors should insist on five answers:
1. What economic problem are we solving? Define the expected improvement in revenue, operating costs, service quality, risk, or customer retention.
2. What must change beyond the technology? Identify the workflows, responsibilities, incentives, and operating procedures that require redesign.
3. How will we measure the result? Establish baseline performance, financial metrics, implementation costs, and clear accountability.
4. What happens to the productivity gains? Determine whether the benefits will support reinvestment, employee development, customer value, stronger margins, or some combination.
5. What evidence would cause us to change course? Establish performance thresholds and review periods before enthusiasm becomes an excuse to continue an underperforming investment.
These questions transform AI from a technology initiative into a leadership discipline.
They also distinguish organizations that are experimenting with the future from those building the capacity to operate successfully within it.
The Future Will Belong to Those Who Deliver
Artificial intelligence may become one of the most consequential technological developments of our lifetime.
Its potential to expand human capabilities, improve economic productivity, and create new opportunities is substantial.
But potential is not performance.
The organizations that thrive will need more than sophisticated models, ambitious spending plans, and impressive demonstrations.
They will need leaders willing to redesign outdated systems, invest in human capabilities, measure results, and accept responsibility for the economic and social consequences of their decisions.
The ultimate measure of this revolution will not be how many tasks machines perform.
It will be how much lasting value people create with the capabilities those machines make possible.
The future will not reward organizations simply for adopting artificial intelligence. It will reward those with the leadership to make intelligence matter.
If you believe leadership should be measured by the opportunities it creates and the results it delivers, subscribe to Global Empowerment Leadership for research-driven perspectives on innovation, economic opportunity, and institutional performance.
Subscribe to Global Empowerment Leadership Subscribe on LinkedIn https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7060440518475804672
Explore how disciplined analysis, strategic investment, and ex*****on help create long-term value at Skyline Property Experts.
The future is being built now. The question is whether we will simply participate in it—or lead it.
Source notes
[1] McKinsey & Company, The State of AI in 2025: Agents, Innovation, and Transformation (2025).
[2] Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond, Generative AI at Work, NBER Working Paper 31161 (2023), subsequently published in The Quarterly Journal of Economics (2025).
[3] Klarna, AI Assistant Handles Two-Thirds of Customer Service Chats in Its First Month, February 27, 2024.
[4] Klarna Group plc, Form 10-K for fiscal year 2025, discussion of AI customer-service operations and reported 2024 savings.
[5] McKinsey & Company, The State of AI: How Organizations Are Rewiring to Capture Value, March 12, 2025.
