For much of modern business history, expertise carried a scarcity premium. Companies paid for specialized knowledge, and leaders built careers on what they knew and how well they could apply it. Entire industries developed around gathering information, analyzing it, and helping others make decisions.
AI is beginning to change that equation. A professional with access to capable tools can now assemble research, draft a proposal, explore strategic options, or develop an initial analysis in minutes. That output still needs checking and judgment; speed is not the same as accuracy or understanding. But the cost and effort of producing a useful first draft are falling.
Adoption is moving quickly. Stanford University's 2026 AI Index, drawing on a McKinsey survey, reports that 88% of respondents said their organizations used AI in at least one business function in 2025, up from 78% in 2024. Those are self-reported survey findings, not a count of every business. Even so, the direction is hard to ignore.
This doesn't make expertise worthless. It changes the question for AI leadership: if more people can produce a competent-looking answer, what makes someone worth listening to, following, or trusting?
What Becomes Scarce When Insights Are Easy to Produce?
My answer is increasingly simple: judgment, credibility, and genuine human connection.
AI can help me research someone before a meeting. It can summarize a company, suggest priorities, and draft thoughtful questions. But it cannot, on its own, make that person trust me. It can produce a persuasive proposal, but it cannot create the history between two executives that makes one willing to take the other's call when something goes wrong.
For nearly three decades, working around enterprise growth and technology transformation, I've watched platforms, tools, and business models change. The opportunities that mattered most in my career rarely came down to knowing a fact no one else knew. They often began because someone trusted me enough to make an introduction, share a problem, or take a chance. That trust had been earned long before it was needed.
Employers still recognize the importance of these capabilities. The World Economic Forum's Future of Jobs Report 2025, based on responses from more than 1,000 employers representing over 14 million workers, identifies leadership and social influence, as well as resilience, flexibility, and agility, among important core skills. It also lists empathy and active listening among the capabilities employers value.
Technology can make information easier to produce. It doesn't remove the need for people who can interpret it responsibly and bring others together to act on it.
Relationship Capital: The Asset Business Often Undervalues
There is a useful name for the value created through these connections: relationship capital. It is an established idea in business research, but one I believe deserves renewed attention in an AI-driven economy.
To me, relationship capital is the trust, credibility, and goodwill accumulated between people over time. It exists between a salesperson and a customer who have worked through difficult periods together. It exists between executives who know that a commitment will be honored. It exists between a leader and a team willing to speak candidly when a plan is failing.
Organizations already measure parts of this through customer retention, employee engagement, referrals, and other indicators. Yet the numbers don't always tell us whether a relationship is strong enough to withstand a difficult conversation or an unexpected setback. A contact list is not the same as trust, and a high customer-satisfaction score doesn't tell the whole story of why someone stays.
These aren't merely sentimental qualities. A trusted colleague may flag a problem sooner. An established partner may help resolve a setback before it becomes a dispute. A customer who believes a business will respond fairly may be more willing to raise a concern directly. Such outcomes are not guaranteed, but anyone who has led a team or managed a long-term account knows the difference relationships can make.
Why AI Leadership Still Depends on Human Trust
Research on teams offers a useful parallel. Google's Project Aristotle studied team effectiveness within Google and identified psychological safety as the most important of five team dynamics in its analysis. In that setting, people on psychologically safe teams felt more able to ask questions, acknowledge mistakes, and take interpersonal risks.
Psychological safety is not identical to relationship capital, and findings from Google cannot automatically be generalized to every workplace. But the research illustrates why the quality of people's interactions matters alongside technical capability.
Consider an employee reviewing an AI-generated recommendation. The system may produce an answer in seconds. Whether the employee challenges an unsupported assumption can depend on the team's culture, the leader's reaction to dissent, and whether anyone has made accountability clear. As Unhyd's reporting on AI workforce skills has explored, expecting people to exercise judgment is not enough if the workflow gives them little room to question the tool.
The same distinction applies to business performance. An AI initiative may deliver apparent returns on investment while important questions about accountability and the quality of decisions remain unanswered. Faster production is useful. The capacity to recognize an error, discuss it openly, and correct course is what helps turn that speed into lasting value.
Low trust creates the opposite conditions. Teams become reluctant to share problems. Decisions require layers of reassurance. Relationships become transactional. An organization can have extraordinary technology and still struggle to coordinate action because the people using it don't trust one another.
Five Questions AI Leaders Should Ask
The strongest organizations will invest in AI. But alongside What can we automate?, I believe leadership teams should ask five questions:
- Where must trust exist for AI adoption to succeed? Identify who will rely on the output, who can question it, and who is accountable when it is wrong.
- Which relationships are we strengthening, and which are we neglecting? Look beyond the number of customer contacts or internal meetings to the quality of difficult conversations and follow-through.
- Where do silos prevent intelligence from becoming action? Make it easier for people across teams to share context and raise concerns before decisions are locked in.
- How do leaders respond when someone challenges an AI-assisted decision? If questioning a recommendation carries a penalty, employees may learn to approve outputs they don't trust.
- Are we measuring outcomes as well as efficiency? Alongside time saved, examine rework, escalations, customer experience, and whether people understand the decisions that affect them.
These questions require practical choices, not another slogan about being human-centered. In a high-stakes workflow, for example, a leader can name the person responsible for reviewing AI output, create a clear route for escalating concerns, and make time for teams to explain what the system missed. Such steps won't replace technical evaluation or sound controls. They make those controls more usable.
The Advantage Is What People Do With the Insight
This is not an argument against AI. I believe organizations should use technology to reduce routine work and extend what capable people can accomplish. The question is whether they are also investing in the relationships and judgment required to use those capabilities well.
When everyone can produce a polished proposal, the proposal itself becomes less distinctive. When many people can generate a plausible insight, the advantage lies in recognizing which insight matters, testing it against reality, and persuading others to act with care. That work depends on expertise, but also on credibility built over time.
I've learned that relationships cannot be summoned at the moment a leader needs support. They develop through consistency, honesty, shared effort, and the willingness to have conversations that aren't always easy.
AI may change how quickly we prepare for those conversations. It does not relieve us of the responsibility to have them. The leaders who stand apart will not simply be those who can generate the best answers, but those who have earned the trust to turn sound judgment into collective action.
