KPMG Partner | Global Data & AI Leader | Enterprise AI, indirect tax compliance, and finance transformation across global organizations.
Over the past two years, I have sat in enough boardrooms across enough industries to notice a pattern. Executives ask me some version of the same question: How much should we trust AI? The framing changes—regulatory risk, operational control, what happens when the model is wrong—but the anxiety is the same.
That question, while understandable, is the wrong one. The more consequential question is this: Are we designing our AI programs to multiply human judgment, or are we just building faster queues?
That is the difference between “human in the loop” and “leverage by design.” Getting it wrong is costing organizations real money, real time and real competitive ground.
The Problem Nobody Talks About
Human in the loop started as a governance idea. Let AI do the heavy lifting, but keep a human in the chain before any final decision is made. Responsible in principle. Sensible on a whiteboard.
In practice, I have watched it become a bottleneck factory.
I have seen organizations deploy machine learning models that classify invoices with 98% accuracy—and then route every single invoice through a human approval queue anyway. The technology is faster. The process is not. The organization automated the creation of work without automating its completion.
This is digital congestion. You have moved the problem, not solved it.
The same pattern shows up in customer service, tax compliance, contract review and supply chain operations. The AI works, but the operating model does not. The people caught in the middle, those reviewing thousands of outputs that were almost certainly correct, are not doing their best work. They are doing the work the machine already did, one more time, more slowly.
That is not a technology failure. It is a design failure.
What ‘Leverage By Design’ Actually Means
The organizations I have seen get this right ask a fundamentally different question. Not where should humans review AI, but where do humans create the most value that AI cannot?
Humans are irreplaceable when situations are ambiguous, when ethical stakes are real, when context outside the data matters or when a judgment call affects a relationship. Those are the moments that define whether a business operates well or merely operates.
Machines are irreplaceable when tasks are repetitive, data is structured, volume is high and rules are clear. Let them work at full speed, without a queue of approvers slowing them down.
This is not an argument for less oversight. It is an argument for better-placed oversight. A finance manager who previously reviewed 500 transactions can now meaningfully oversee 50,000, if the system is designed to surface what actually needs attention. A tax professional who once spot-checked samples can monitor entire populations for anomalies. That is leverage, and it requires intentional architecture.
The Trust Question Deserves A Better Answer
Organizations default to reviewing everything because they confuse review with trust—assuming a human eye on every output makes the system trustworthy. Trust comes from system design, not volume of review.
A well-governed AI process should answer five questions clearly: What decision is being automated? What confidence level is required before acting? What triggers an escalation to a human? Who is accountable for the outcome? How is performance measured on an ongoing basis?
When those questions have clear answers, you do not need a person approving every invoice. You need a person who owns the system, watches the signals and intervenes when something unusual happens.
Aviation figured this out decades ago. Modern aircraft are extensively automated. Pilots do not manually fly every phase of every flight. They monitor, interpret and take control when the situation calls for it. No passenger boards thinking automation makes the flight riskier. The design is intentional and the governance is continuous.
Enterprises are not there yet, but they can be.
Four Moves That Actually Work
The path from human in the loop to leverage by design follows four moves.
• Automate the high-volume, low-ambiguity work entirely. Where rules are clear, data is structured and errors are detectable, stop reviewing manually.
• Redesign human intervention around exceptions, not transactions. Humans engage when something unusual happens, not as a default checkpoint on everything that happened.
• Build real-time performance monitoring into the system from day one. Not quarterly audits. Continuous metrics that tell you whether the system is performing as designed and where it is drifting.
• Move your best people upstream. Stop using senior talent to review AI outputs. Use them to improve the systems, own accountability for outcomes and handle situations that genuinely require human judgment. That is where the organization wins.
The Leadership Question This Raises
None of this is a technology problem. The technology is ready. What is not ready is the decision architecture—which decisions belong to humans, which to machines and which require both working together.
Consider a global manufacturer that automated routine procurement approvals across its supply chain. Rather than adding reviewers, it redirected operations leaders toward supplier risk and strategic sourcing. A healthcare network made a similar move by automating claims triage at scale while keeping clinical staff focused on complex case decisions. Both gained scale without losing accountability because leadership defined the architecture deliberately.
This is becoming a board-level conversation. Not operational. Strategic. How much autonomy does your AI have? Where does accountability sit when a machine-generated recommendation turns out to be wrong? What does transparency look like for regulators, auditors and customers?
Organizations that answer these questions deliberately will scale faster than those still debating whether AI belongs in critical workflows at all.
The goal was never autonomy for its own sake. The goal is an organization where people spend less time processing and more time thinking. That is leverage by design. And, in my experience, it is where the real return on AI investment begins.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?