~7-minute read

Editor’s Note

In Part V, Decision Boundaries, I explored a question that becomes increasingly important as AI moves from recommendation into execution:

Under what conditions should decision authority shift?

But once some authority does shift, another question follows:

What becomes the leader’s job?

Traditional operating models distribute responsibility through layers of human coordination.

Leaders set intent.

Managers translate intent into priorities.

Teams execute and make operating decisions.

Exceptions escalate.

Leaders review outcomes and adapt.

AI increasingly participates in the middle of that chain.

It can coordinate workflows.

Recommend actions.

Make bounded decisions.

Execute approved actions.

Monitor outcomes.

That does not make leadership less important.

It changes where leadership creates value.

What Leaders Need to Know

Intent becomes more important.

AI can optimize execution, but leaders still determine the outcomes, priorities, constraints, and trade-offs that matter.

Authority must be designed.

Delegation requires explicit boundaries, thresholds, controls, evidence, and escalation conditions.

Oversight shifts from activity toward system behavior.

Leaders need visibility into outcomes, exceptions, overrides, and emerging patterns — not necessarily every individual action.

Human judgment should remain closest to decisions where ambiguity and consequence are highest.

The leadership challenge is therefore not simply to supervise more capable technology.

It is to redesign the operating environment around it.

The Leadership Leverage Framework™

If intelligent systems increasingly participate in execution, where should leadership concentrate its attention?

I see four responsibilities becoming increasingly important:

Set Intent → Design Boundaries → Observe the System → Intervene & Adapt

At the center sits:

Accountability for Outcomes

The four responsibilities form a continuous leadership loop.

Leaders define the intent.

They determine the authority boundaries within which people and AI operate.

They observe how the system behaves against intended outcomes.

And they intervene when judgment is required — adapting the objective, boundary, control, or operating model as conditions change.

Accountability sits at the center because delegation does not remove responsibility for outcomes.

The objective is not to keep leadership involved in every action.

It is to concentrate leadership where it creates the greatest leverage.

The underlying principle is:

Leadership moves upstream — from managing individual actions to designing the conditions under which good decisions happen.

And as AI gains more autonomy:

Leadership does not disappear. Its leverage point moves.

From Managing Actions to Designing the System

When humans perform most of the work, leadership naturally sits close to execution.

A proposal is reviewed.

A decision is authorized.

An exception is escalated.

An outcome is checked.

That model works when the number and speed of decisions remain manageable.

But imagine an enterprise where hundreds of intelligent systems participate in thousands of operational decisions.

If every meaningful action still waits for a leader, the organization may increase AI capability without increasing organizational speed.

The leadership question therefore changes.

Instead of asking:

Which decisions should I personally approve?

leaders may increasingly need to ask:

What conditions should exist so the right decisions can happen without requiring me to approve every one?

That is a different leadership discipline.

From Hierarchy to Leadership Architecture

Traditional organizations often distribute authority through hierarchy.

When uncertainty appears, a question moves upward until it reaches someone with sufficient authority.

AI creates another possibility.

Some authority can increasingly be designed into the operating environment through permissions, thresholds, decision boundaries, execution controls, escalation rules, and observability.

This does not eliminate hierarchy.

But it can reduce the need for hierarchy to participate in every transaction.

Leadership therefore begins moving from being the decision point toward designing the decision architecture.

That distinction matters.

A leader who personally resolves every exception may appear highly involved.

But a leader who designs a system that handles routine situations appropriately — while surfacing the exceptions requiring human judgment — may create far greater organizational leverage.

1. Set Intent

AI systems can optimize.

But leaders still determine what should be optimized for.

Speed?

Customer experience?

Cost?

Resilience?

Risk reduction?

Growth?

Those objectives frequently compete.

A system can execute flawlessly against the wrong objective if organizational intent is unclear.

Leadership therefore begins by defining:

the outcome

the priority

the constraints

the acceptable trade-offs

and what must not be optimized away

The more execution becomes distributed, the more important clarity of intent becomes.

Setting intent is no longer simply a communication exercise.

It becomes part of the operating architecture.

2. Design Boundaries

Intent tells the organization where to go.

Boundaries determine how far AI is allowed to go.

Which decisions may AI make?

Which should it recommend?

Which require approval?

When must the system escalate?

Part V focused on these Decision Boundaries.

The core principle was:

The matrix classifies the decision.
The boundary determines who gets to act.

That now becomes part of the leader’s operating responsibility.

Delegation to AI should not happen simply because the technology is capable.

Capability answers what AI can do.

Authority determines what it is allowed to do.

And one governance principle remains important:

No delegated authority = no autonomous action.

3. Observe the System

Traditional supervision often focuses on individual activity.

Who did what?

Was the task completed?

Was the process followed?

As execution becomes increasingly distributed across humans and AI, leaders may need to move from watching individual actions toward observing system behavior.

Are outcomes improving?

Where are exceptions increasing?

Where are people repeatedly overriding AI recommendations?

Which boundaries are being reached most frequently?

Are systems optimizing locally while creating enterprise-level problems?

Are outcomes moving away from organizational intent?

The objective is not to watch every decision.

It is to remain able to see when the system begins moving in the wrong direction.

In an AI-native enterprise, observability is therefore not only a technology capability.

It increasingly becomes a leadership capability.

4. Intervene & Adapt

Human judgment does not disappear.

It becomes concentrated where it matters most.

Novel situations.

Conflicting objectives.

High-consequence decisions.

Ethical trade-offs.

Strategic ambiguity.

Unexpected system behavior.

Leadership therefore becomes partly about knowing when to enter the loop.

Sometimes the right intervention is to override a decision.

But often the more valuable intervention is to:

redesign the boundary

change the objective

modify the control

adjust the threshold

or redesign the operating model

That is the difference between fixing one decision and improving the system that produces decisions.

And the loop then begins again.

New evidence may require new intent.

Changed conditions may require new boundaries.

Observed behavior may require intervention.

Leadership therefore becomes continuous and adaptive.

Accountability Does Not Require Constant Approval

This may be one of the hardest leadership shifts.

Organizations often equate accountability with approval.

If a leader is accountable, the assumption is that the leader should personally approve the action.

But those are different concepts.

Approval is participation in a specific decision.

Accountability is responsibility for the outcomes, controls, and operating environment in which decisions occur.

A leader can remain accountable for outcomes while delegating execution — provided intent, authority, controls, evidence, and escalation are clearly designed.

This distinction matters because human approval does not scale indefinitely.

As AI participates in more decisions, keeping leaders in every loop can recreate the coordination bottleneck the technology was intended to remove.

The question becomes:

How do we preserve accountability without requiring constant participation?

That is why Accountability for Outcomes sits at the center of the Leadership Leverage Framework™.

It is not another stage.

It is the enduring responsibility connecting all four.

Three Leadership Failure Modes

The transition can fail in predictable ways.

Over-supervision

AI gains capability, but every meaningful action still requires human approval.

The technology changes.

The operating model does not.

Over-delegation

Authority expands faster than evidence, controls, or accountability.

The organization gains speed but loses confidence.

Under-observation

Leaders step away from individual actions without building mechanisms to understand system-level behavior.

Autonomy increases.

Visibility decreases.

The objective is not to choose between autonomy and control.

It is to design an operating model in which both can scale together.

Consider Major Incident Management

Imagine a critical enterprise service disruption.

AI detects degradation.

It correlates signals across systems.

It identifies probable causes.

It opens collaboration channels.

It runs approved diagnostics.

It recommends recovery actions.

And within predefined boundaries, it may execute certain remediation steps.

Where does leadership create value?

Not by approving every diagnostic action.

Leadership creates value by determining:

What business outcome takes priority?

What level of risk is acceptable?

Which remediation actions may proceed within defined boundaries?

What requires human authorization?

When should customer impact outweigh technical optimization?

When has the situation moved outside the original decision boundary?

When should the response strategy change entirely?

AI may increasingly manage parts of execution.

Leadership remains responsible for defining the operating context.

That is the distinction.

From Management Presence to Management Leverage

In traditional operating models, leadership presence is often visible through participation.

The leader attends the meeting.

Reviews the proposal.

Approves the request.

Resolves the escalation.

Signs off on the outcome.

In an AI-native operating model, some of those visible acts may occur less frequently.

But reduced participation does not necessarily mean reduced leadership.

The leader may have created greater leverage upstream.

By defining the objective.

By designing the authority boundary.

By establishing evidence requirements.

By making exceptions observable.

By determining when escalation is required.

By ensuring the system can be corrected when conditions change.

This is why leadership in an AI-native enterprise may become less about being present in every decision and more about shaping the conditions under which decisions are made.

The Leadership Shift

The traditional center of gravity has often been:

Direct.
Decide.
Approve.
Review.

The AI-native leadership model may increasingly shift toward:

Set intent.
Design boundaries.
Observe the system.
Intervene and adapt.

This does not make leaders passive.

The work becomes more architectural.

Leadership increasingly shapes the environment in which decisions occur rather than participating in every decision itself.

A Question for the Leadership Team

Imagine that intelligent systems eventually participate in a meaningful share of enterprise execution.

Which leadership activities should remain fundamentally human?

Which should become increasingly automated?

Which should move from direct participation toward oversight?

Where should authority remain fixed?

Where should it expand as evidence improves?

And which new leadership responsibilities will emerge because intelligent systems are participating in execution?

Those are not primarily technology questions.

They are organizational design questions.

Leadership teams may need to answer them before technology answers them by default.

Final Thought

AI-native leadership is not about stepping out of the loop.

It is about becoming more deliberate about where leadership should enter it.

AI can increasingly coordinate.

AI can recommend.

AI can act.

Within defined boundaries, AI can increasingly decide.

But leaders still determine:

what the organization is trying to achieve

how much authority should be delegated

what evidence is sufficient

where the limits are

what outcomes matter

and when human judgment must intervene

Leadership does not disappear as AI gains autonomy.

Its leverage point moves.

The shift may look like this:

From supervising activity
to designing operating conditions.

From approving decisions
to defining decision environments.

From managing individual actions
to governing outcomes.

From being in every loop
to knowing when judgment matters most.

So where should human leadership create the greatest leverage?

Where intent, trade-offs, boundaries, accountability, and judgment must be defined — not in routine execution itself.

The greatest leadership leverage will not come from being in every decision.

Continue the Series

Season 5 · Leading the AI-Native Enterprise

Part I — The AI-Native Enterprise
How enterprise transformation changes when AI becomes part of the operating model.

Part II — The Enterprise Coordination Shift™
How AI changes how work moves across the enterprise.

Part III — Enterprise Trust Architecture™
How autonomy can scale without losing control.

Part IV — Human + AI Execution Models
How responsibility can shift dynamically between humans and intelligent systems.

Part V — Decision Boundaries
How enterprises determine where AI should decide, recommend, wait for approval, or escalate.

Part VI — Leadership in the AI-Native Enterprise
How leadership changes when intelligent systems increasingly participate in execution.

AI at the Frontlines

Leadership · Transformation · Impact

Ideas. Frameworks. Insights. Impact.

Editor’s Note: In Part IV, Human + AI Execution Models, I explored how responsibility can shift dynamically between humans and intelligent systems across the execution cycle.

But once AI is allowed to act, another question becomes unavoidable:

Who decides what AI gets to decide?

AI systems are becoming capable of doing more than generating answers.

They can interpret context, recommend actions, coordinate workflows — and increasingly, make decisions and act on them.

But capability creates a different management question:

Just because AI can make a decision, should it be allowed to?

That distinction matters.

Capability answers what AI can do.
Decision rights determine what AI is allowed to decide.

What Leaders Need to Know

Decision context comes first. Ambiguity and risk help determine what kind of decision is being made.

Authority comes next. AI may decide, recommend, wait for approval, or escalate.

Human oversight should be intentional friction. Decision boundaries can expand — or contract — as evidence grows.

Start With the Decision Context

Recent MIT CISR research offers a useful way to simplify the problem.

Its AI Decision Matrix evaluates decisions using two dimensions:

Ambiguity — how clearly the available information determines the answer.

Risk — what happens if the decision is wrong.

Together, they create four decision contexts:

Routine — low ambiguity, low risk
Exploratory — high ambiguity, low risk
Consequential — low ambiguity, high risk
Strategic — high ambiguity, high risk

The key insight is simple:

Different decision contexts require different combinations of human and AI participation.

MIT’s research notes that routine decisions are stronger candidates for automation, while strategic decisions require stronger human leadership and oversight. Human involvement remains important across the categories, but the role changes.

Figure 1. Decision Landscape — Four types of decisions, defined by ambiguity and risk. Based on MIT CISR’s AI Decision Matrix.

That framing moves the conversation from:

“Is AI capable enough?”

toward:

“What kind of decision are we asking AI to make?”

But enterprise execution requires one more step.

Set the Decision Boundary

Once the decision context is understood, leaders still need to determine:

What authority should AI actually receive?

Four operating postures help translate decision context into execution authority.

DECIDE

AI acts within explicitly delegated boundaries.

RECOMMEND

AI analyzes and proposes; humans frame, judge, and choose.

APPROVE

AI may prepare or initiate the action, but an authorized human approves execution.

ESCALATE

AI stops and transfers responsibility when the situation exceeds its boundaries.

The matrix classifies the decision.
The boundary determines who gets to act.

Figure 2. Decision Boundary Overlay — Typical operating posture after the decision context is understood.

These are not rigid mappings to the four MIT quadrants.

They are boundaries enterprises design based on context, controls, evidence, and risk tolerance.

A routine decision may begin in recommendation mode while a new agent is being validated.

A consequential decision may allow automation for some actions but require human approval for others.

A strategic decision may use AI heavily for analysis while keeping framing and final authority firmly human.

And beneath the entire model sits one non-negotiable rule:

No delegated authority = no autonomous action.

Technical capability does not create organizational permission.

Authority has to be deliberately granted through policy, permissions, architecture, operating rules, and accountability.

This connects directly to the Enterprise Trust Architecture™ from Part III.

Decision boundaries need to be enforceable, not merely documented.

Human Approval Is Not Free

Human approval is often treated as a safeguard by default.

But every approval gate introduces latency.

Paul Cheek, in No One Works Here, argues that organizations can create human bottlenecks when people remain in the loop for decisions machines could make faster. He also describes a temporary “reliability tax”: when an agent is first deployed, organizations may spend more time validating its output before they have enough evidence to expand its autonomy.

Cheek’s Ferrari-engine-in-a-horse-drawn-carriage metaphor captures the broader problem well: adding powerful AI to an unchanged operating model may leave the enterprise constrained by the old process.

That creates a useful management test:

What risk does human approval reduce — and what delay does it introduce?

Sometimes slowing the system is exactly the right control.

But in other situations, delay creates its own cost — slower customer response, missed windows, longer recovery, or simply a new approval bottleneck.

Human approval should be intentional friction — not default friction.

The Reliability Tax in Practice

Consider an AI agent handling routine customer service credits.

The policy is clear.

The customer facts are available.

The financial exposure is limited.

The agent may already be technically capable of deciding whether a small credit fits policy.

But when the system is first deployed, the organization may still require:

AI recommends → Human reviews → Human approves → Action executes

At first, the AI-enabled process may actually be slower.

That does not necessarily mean the system has failed.

It may mean the organization is paying the reliability tax — spending extra time validating the agent before increasing its authority.

The important question is what happens next.

As evidence accumulates, leaders can monitor accuracy, override rates, policy exceptions, customer outcomes, fraud or leakage, and time to resolution.

If reliability remains strong, the boundary can move from:

AI recommends → Human approves

to:

AI decides within defined limits → Humans monitor exceptions

And if reliability deteriorates, the boundary can contract again.

Autonomy should be earned through evidence, not assumed through capability.

A useful progression is:

Start with validation.
Earn confidence through evidence.
Expand autonomy within boundaries.
Contract the boundary when reliability falls.

This makes autonomy dynamic rather than binary.

The question is no longer:

“Should humans stay in the loop?”

It becomes:

What evidence should AI have to earn before the loop can change?

Boundaries Should Move With the Decision

Decision boundaries are not simply labels assigned once to a process.

They can shift as the situation changes.

Take the same customer-service agent operating under a defined autonomous-credit limit.

A straightforward request with complete information may remain inside the AI Decides boundary.

A request with conflicting account information may move into AI Recommends.

A larger credit may require Human Approval.

A suspected fraud pattern may trigger Escalation.

What changed was the decision context — and therefore the authority boundary.

That is the practical connection between Part IV and Part V.

Part IV asked:

Where should human judgment enter the execution loop?

Part V asks:

Under what conditions should decision authority shift?

The Leadership Question

Organizations have always designed authority.

Employees have spending limits.

Managers have approval thresholds.

Operational teams have escalation paths.

AI does not remove that responsibility.

It expands it.

Leaders may increasingly become accountable not only for individual decisions, but for designing the boundaries within which intelligent systems are allowed to decide.

That changes the management question from:

“Did I approve this decision?”

to:

“Did we design the right decision boundary?”

That is a fundamentally different leadership responsibility.

The objective is not maximum autonomy.

It is appropriate authority.

Final Thought

The AI-native enterprise will not scale by giving AI unlimited decision authority.

Nor will it scale if every meaningful action eventually waits for a person.

The more durable model sits between those extremes.

Understand the decision context.

Set the authority boundary.

Validate the system.

Expand autonomy when evidence supports it.

Contract the boundary when conditions change.

Because once AI can act, the defining question is no longer simply:

What can AI do?

It becomes:

Who decides what AI gets to decide?

Continue the Series

This is Part V of Season 5 — Leading the AI-Native Enterprise.

Part I — The AI-Native Enterprise
Why enterprise transformation changes when AI becomes part of the operating model.

Part II — The Enterprise Coordination Shift™
How AI changes how work moves across the enterprise.

Part III — Enterprise Trust Architecture™
How autonomy can scale without losing control.

Part IV — Human + AI Execution Models
How responsibility can shift dynamically between humans and intelligent systems.

Part V — Decision Boundaries
How enterprises determine where AI should decide, recommend, wait for approval, or escalate.

Coming Next — Part VI: Leadership in the AI-Native Enterprise

When intelligent systems increasingly coordinate work, execute actions, and make decisions within defined boundaries:

What becomes the leader’s job?

AI at the Frontlines
Leadership · Transformation · Impact

Ideas. Frameworks. Insights. Impact.

Research Note

The decision-context framing above draws on the AI Decision Matrix developed by Ina M. Sebastian, Peter Weill, Thomas Haskamp, and Jan vom Brocke at MIT CISR and featured by MIT Sloan on September 22, 2026.

The discussion of the reliability tax draws on Paul Cheek’s No One Works Here, excerpted by MIT Sloan.

The Ferrari-engine / horse-drawn-carriage analogy is attributed to Cheek in a U.S. Chamber of Commerce discussion of AI transformation and workflow redesign.

~6-minute read

Editor’s Note: In Part III, Trust Is Becoming Infrastructure, I argued that increasing AI autonomy requires trust to be designed into the enterprise through authority, boundaries, controls, observability, and accountability.

But once those foundations exist, another question emerges:

How should humans and intelligent systems actually share responsibility for execution?

Imagine an enterprise running hundreds of AI agents across IT, finance, procurement, and customer operations.

Each can interpret events, coordinate workflows, make routine decisions, and take action within defined boundaries.

Now imagine requiring a human to approve every consequential action.

The organization has created a safeguard.

It may also have created its next bottleneck.

Human-in-the-loop has become an important principle of responsible AI.

And rightly so.

But as AI moves from assistance toward execution, enterprises may need to separate two ideas that are often treated as synonymous:

Human oversight and human approval.

The challenge is not removing humans from enterprise execution.

It is determining where human judgment belongs within it.

That leads to a broader management question:

How should responsibility for execution be distributed between humans and intelligent systems?

This may become one of the defining operating-model questions of the AI-native enterprise.

When the Safeguard Becomes the Bottleneck

Much of enterprise AI still follows a familiar pattern:

AI identifies → AI recommends → Human reviews → Human approves → System executes

For relatively low volumes of AI-mediated decisions, this model works well.

But now consider an organization where intelligent systems continuously monitor events, interpret information, coordinate workflows, and initiate actions across multiple business functions.

An AI agent detects an operational anomaly.

Another interprets its potential impact.

Another identifies possible actions.

A workflow is initiated.

A decision is prepared.

And then everything stops.

A human must approve what happens next.

Now multiply that pattern across hundreds or eventually thousands of AI-mediated actions.

If every meaningful action eventually enters a human approval queue, the enterprise has not eliminated the coordination bottleneck.

It has moved it to the human approval layer.

The answer is not unrestricted autonomy.

Enterprise execution has always operated through different levels of delegated authority.

Employees have spending limits.

Managers have approval thresholds.

Applications have permissions.

Operational teams have escalation procedures.

AI systems will need similar boundaries.

An AI agent might resolve a routine IT request independently but escalate a production change.

It might execute a transaction below a defined threshold but require authorization above it.

It might communicate independently in a routine customer interaction but escalate when financial, regulatory, or reputational consequences increase.

The objective is neither maximum autonomy nor maximum human control.

It is appropriate autonomy.

And that requires a more deliberate model for how humans and AI participate in execution.

The Human–AI Execution Loop™

Enterprise work rarely consists of one isolated decision.

Organizations continuously observe conditions, interpret information, decide, act, and learn from outcomes.

As AI becomes more capable, it can increasingly participate across that entire cycle.

I think about this through five stages:

Sense → Interpret → Decide → Act → Learn & Govern ↻

1. Sense

Every execution cycle begins with signals.

A customer request arrives.

A service fails.

Inventory falls below a threshold.

A financial anomaly appears.

A project dependency changes.

An emerging operational risk is detected.

AI can continuously monitor these signals at a scale and speed humans cannot realistically replicate.

But humans still determine an important question:

What matters enough to monitor?

What constitutes risk?

Which signals require attention?

Which outcomes matter to the enterprise?

AI may dramatically expand the organization’s ability to sense.

Humans still shape what the organization chooses to care about.

2. Interpret

Signals require context.

What happened?

Why might it have happened?

What other information matters?

What could happen next?

AI can retrieve enterprise context, combine information, identify patterns, and generate possible interpretations.

But interpretation becomes more difficult when situations involve ambiguity, competing objectives, organizational nuance, incomplete information, or significant consequences.

This is where human judgment may remain particularly valuable.

The question is not whether AI can interpret information.

It is whether the interpretation is reliable enough for the consequence of the decision that follows.

3. Decide

Interpretation eventually leads to a decision.

But not every decision requires the same decision-maker.

Depending on the situation:

AI recommends → Human decides

or

AI decides → Human approves

or increasingly,

AI decides within defined boundaries

The important question is not simply whether AI can make a decision.

It is:

Under what conditions should the enterprise allow it to do so?

That distinction becomes increasingly important as AI moves closer to execution.

4. Act

A decision creates value only when something happens.

An AI system might update a record.

Trigger a workflow.

Communicate with a customer.

Modify a configuration.

Complete a transaction.

Coordinate another agent.

Or initiate a sequence of actions across multiple enterprise systems.

This is where recommendation and execution diverge.

A poor recommendation can be rejected.

An incorrect autonomous action may already have consequences.

Execution therefore requires authority, enforceable controls, observability, and escalation paths.

The question is no longer only:

Was the AI correct?

It also becomes:

Was the AI authorized to act?

5. Learn & Govern

Execution produces outcomes.

And outcomes produce evidence.

What worked?

What failed?

Where did humans intervene?

Which exceptions occurred?

Which exceptions repeatedly occurred?

Were the boundaries too restrictive?

Were they too permissive?

Should authority change?

This is where human responsibility becomes particularly important.

The objective is not simply to improve the AI system.

It is to improve the execution system around it.

And then the cycle begins again:

Sense → Interpret → Decide → Act → Learn & Govern → Sense

The More Important Question: Who Owns What?

The execution loop alone is not the framework’s most important idea.

The more consequential management question is:

What level of responsibility should humans and AI carry at each stage, under which conditions?

I see four useful responsibility modes.

Human-led

Humans own the activity while AI provides information or support.

This remains appropriate where judgment, ambiguity, relationships, ethics, or consequences dominate.

AI may inform the work.

But the human retains responsibility for deciding and acting.

AI-assisted

AI analyzes, recommends, summarizes, predicts, or prepares an action.

A human retains consequential decision authority.

This is where much of enterprise AI operates today.

AI increases the quality or speed of human execution without fundamentally changing who holds authority.

AI-delegated

AI receives authority to decide and act within explicitly defined boundaries.

Humans establish those boundaries.

AI operates inside them.

Exceptions move back to humans.

This is where enterprises begin moving from AI assistance toward meaningful AI-enabled execution.

AI-autonomous with human oversight

AI continuously operates within an established environment while humans monitor outcomes, exceptions, patterns, and systemic behavior.

Humans no longer approve every transaction.

But they remain accountable for the system.

And this is the important point:

These modes can coexist within the same process.

An AI agent might autonomously handle routine cases.

It might request human approval when a threshold is crossed.

It might transfer control entirely to a human when uncertainty increases.

And it might return to autonomous operation once the exception has been resolved.

Human + AI execution is therefore not simply a handoff.

It is a dynamic allocation of responsibility.

From Human-in-the-Loop to Human-on-the-Loop

This creates an important distinction.

With human-in-the-loop, the human participates directly in the execution path.

AI recommends → Human reviews → Human approves → Execution continues

For consequential, ambiguous, or uncertain decisions, that may remain exactly the right model.

But higher levels of trusted autonomy create another possibility:

Human-on-the-loop.

In this model, humans establish authority and boundaries.

AI operates within them.

The environment remains observable.

Exceptions trigger intervention.

Humans monitor outcomes and adjust the system when necessary.

The human begins moving from being a transactional approver toward becoming a system-level supervisor.

Human-in-the-Loop vs Human-on-the-Loop — The difference between transaction-level approval and system-level oversight.

That does not mean humans become less important.

In some ways, their responsibility becomes greater.

A manager approving an individual transaction is responsible for that decision.

A leader defining the environment within which thousands of AI-mediated decisions occur is responsible for something broader:

The design of the decision environment itself.

This should not be interpreted as a maturity ladder where human-on-the-loop is always superior.

It isn’t.

The appropriate position of the human depends on factors such as consequence, uncertainty, and delegated authority.

Some activities should remain human-led.

Some require humans directly in the loop.

Others may operate effectively with humans supervising the loop.

The goal is not less human involvement.

It is better-placed human judgment.

Which leads to the central proposition:

Human oversight does not necessarily mean human approval.

The Management Challenge Is Execution Design

Discussions about humans and AI often focus on jobs and tasks.

Which tasks will AI automate?

Which skills will employees need?

Which roles will change?

Those questions matter.

But AI capable of taking action introduces another management problem:

How should the enterprise distribute execution responsibility between humans and intelligent systems?

That is not simply a technology question.

It is an operating-model question.

Business leaders need to determine where human judgment creates disproportionate value.

Technology leaders need to make delegated authority technically enforceable.

Risk leaders need to define consequences, thresholds, and escalation conditions.

Product and process owners need to determine when AI should recommend, decide, execute, or escalate.

And leadership needs to ensure accountability remains clear even when humans no longer approve every individual action.

This is why Human + AI Execution cannot simply mean inserting AI into an existing workflow.

The workflow itself may need to be redesigned.

An Executive Test

Take one enterprise process where AI could increasingly take action.

It could be an IT incident.

A procurement request.

A customer interaction.

A financial exception.

A supply-chain disruption.

A security event.

Now ask five questions:

1. What can AI sense that humans currently monitor?

2. What can AI interpret reliably enough to support execution?

3. Which decisions could be delegated within defined boundaries?

4. Which actions could AI safely execute?

5. Where must human judgment, oversight, or accountability remain?

The result probably will not be a simple division between “human work” and “AI work.”

It is more likely to produce a portfolio of responsibility.

Some activities remain human-led.

Some become AI-assisted.

Some are delegated.

Others may eventually operate autonomously under human oversight.

And the appropriate mode may change as risk, uncertainty, context, and consequence change.

That portfolio becomes the foundation of a Human + AI Execution Model.

The Next Question

But dynamic responsibility creates another challenge.

If humans and AI can assume different roles across the execution loop:

Who determines the boundaries?

Which decisions can AI make independently?

Which should AI recommend?

Which require human approval?

Which should always be escalated?

What level of uncertainty is acceptable?

What happens when consequences increase?

And what causes a decision to move from one category to another?

Those questions take us from execution models to decision rights.

And that is where the next chapter begins.

Final Thought

The AI-native enterprise will not be defined simply by how many AI tools its employees use.

Nor will it be defined by removing humans from execution.

It will be defined by something more nuanced:

How intelligently it distributes responsibility between humans and intelligent systems.

Sometimes humans will lead.

Sometimes AI will assist.

Sometimes authority will be delegated.

Sometimes AI will operate while humans supervise the system around it.

The objective is not to keep humans involved in every action.

It is to ensure human judgment enters where it matters most.

Because as AI moves from assistance toward execution:

Human oversight does not necessarily mean human approval.

Increasingly, it may mean designing the boundaries within which intelligent systems can act responsibly.

Continue the Series

This is Part IV of Season 5 — Leading the AI-Native Enterprise from AI at the Frontlines.

Part I — The AI-Native Enterprise
Why enterprise transformation changes when AI becomes part of the operating model.

Part II — The Enterprise Coordination Shift™
How AI may change coordination more profoundly than individual tasks.

Part III — Enterprise Trust Architecture™
How enterprises can support increasing autonomy without losing control.

Part IV — Human + AI Execution Models
How responsibility for execution can be distributed between humans and intelligent systems.

Coming Next: Part V — Decision Boundaries

Which decisions should AI make, recommend, or escalate?

AI at the Frontlines explores the management architecture of the AI-native enterprise — how coordination, trust, execution, decision rights, leadership, and operating models change as intelligent systems take on greater responsibility.

Stay at the Frontlines

Subscribe for free to receive new research, frameworks, and executive briefings on the transition to the AI-native enterprise. No paid subscription required.

~7-minute read

Editor’s Note: In Part II, The Enterprise Coordination Shift™, I explored what happens when AI begins to coordinate work rather than simply complete tasks. That shift raises another question: How do enterprises build confidence in systems that increasingly participate in decisions and execution?

Part III explores why trust may need to evolve from a governance principle into part of the enterprise architecture itself.

Every enterprise runs on trust.

We trust people with information, managers with decisions, systems with transactions, and processes with consistency. Where trust has limits, organizations build controls around it.

For decades, those controls were designed around a relatively stable assumption:

People make decisions. Systems execute them.

Agentic AI begins to challenge that assumption.

AI systems are increasingly capable of interpreting context, coordinating workflows, interacting with other systems, recommending decisions and, within defined conditions, taking action.

The question for leaders therefore changes.

It is no longer simply:

Can we trust the AI’s answer?

Increasingly, it becomes:

Can we trust the system to act?

That is a fundamentally different management problem.

As AI becomes part of enterprise execution, trust cannot depend entirely on humans reviewing every action.

Trust may need to become infrastructure.

The Trust Paradox

Consider how enterprise AI is evolving.

A copilot drafts an email. A human reviews it.

An AI system recommends the next action in a workflow. A human approves it.

An AI agent interprets a situation, coordinates across systems and executes an action within predefined authority.

Each step creates more potential value. But each step also changes the nature of trust.

When AI generates content, trust largely concerns the quality of the output.

When AI recommends a decision, trust expands to include context and consequences.

When AI acts, trust extends into authority, execution and accountability.

Enterprises therefore face a paradox.

The simplest way to create confidence is to place a human behind every consequential AI action.

But if every meaningful action requires human approval, much of the value of autonomy disappears.

At the other extreme, giving AI systems broad autonomy without sufficient controls creates operational, regulatory and reputational risk.

The challenge is not choosing between autonomy and control.

It is designing autonomy with control.

Governance Is Necessary. But It May Not Be Enough.

Most large organizations are already developing AI governance capabilities.

Responsible AI principles are being defined. Risk classifications are being established. Models are being reviewed. Policies are being written. Oversight structures are being created.

All of this matters.

But increasingly autonomous systems introduce another challenge.

A policy can define what an AI system should do.

The operating architecture determines what that system can do.

Consider a policy stating that an AI agent cannot authorize a financial transaction above a defined threshold.

That is governance.

But what technically prevents the agent from initiating the transaction?

What happens when confidence falls below an acceptable level?

What if two agents attempt conflicting actions?

When must a human be brought into the process?

And can the enterprise reconstruct what happened afterward?

These questions cannot be answered by policy alone.

They require controls embedded into the environment in which intelligent systems operate.

This is why I believe the conversation will increasingly expand from AI governance toward a broader concept:

Enterprise Trust Architecture™

Enterprise Trust Architecture™ is a way of designing authority, context, decision boundaries, execution controls, observability and human oversight so intelligent systems can operate with increasing autonomy while remaining aligned with enterprise intent.

In this model, trust is not assumed.

It is designed.

Six Elements of Enterprise Trust

I see six interconnected elements becoming increasingly important as AI moves deeper into enterprise execution.

1. Identity & Authority

Before an AI system can act, the enterprise needs to know what it is and what authority it has.

Who owns the agent? Which systems can it access? What data can it use? What actions can it perform?

Enterprises already manage authority for employees, applications and services. Agentic systems extend that requirement to a new class of enterprise actor.

Trust starts by knowing who—or what—is acting, and under whose authority.

2. Context & Grounding

Authority alone does not make a decision trustworthy.

An intelligent system also needs reliable context.

What information is it using? Is it current? Which enterprise source is authoritative? What assumptions is the system making?

An AI system can reason effectively and still reach the wrong conclusion if its context is incomplete, outdated or conflicting.

Trust therefore depends not only on model capability, but on context integrity.

3. Decision Boundaries

Not every decision deserves the same degree of autonomy.

Some decisions may be safe for AI to make independently. Others may require human approval. Some may be delegated only within financial, operational, regulatory or confidence thresholds.

And some decisions should remain entirely human.

The key question becomes:

What is this system allowed to decide under these conditions?

Autonomy should not be an on/off switch.

It needs boundaries.

4. Execution Controls

A trustworthy recommendation and a trustworthy action are different things.

Once an AI system can act, enterprises need to govern execution itself.

Which systems can it modify? Which transactions can it initiate? Which actions require additional authorization? Can an action be stopped or reversed?

The closer AI moves to execution, the more controls need to exist inside the execution environment, not simply around the model.

5. Observability & Evidence

Enterprises cannot govern what they cannot see.

Traditional technology observability asks whether a system is available, performing correctly or generating errors.

AI-native observability will need to answer different questions.

What decision was made? What context influenced it? Which agents and systems participated? What actions followed? Where did a human intervene? What was the resulting business outcome?

The objective is not to record every internal computation.

It is to create sufficient evidence to understand and govern consequential AI-mediated execution.

Without observability, autonomy can become opacity.

6. Human Oversight & Accountability

Human judgment remains essential.

But where humans participate may change.

Today, organizations frequently create confidence by placing humans in the loop, reviewing and approving individual actions.

As the volume and speed of AI-mediated decisions increase, that model becomes difficult to scale.

Humans may increasingly move onto the loop: defining boundaries, monitoring outcomes, reviewing exceptions and intervening when thresholds are crossed.

The objective is neither maximum AI autonomy nor maximum human control.

It is appropriate autonomy with clear accountability.

The Scalability Problem

This distinction matters because human approval has a hidden cost.

Every approval requires someone to receive a request, understand the context, evaluate the recommendation, make a decision and potentially document or escalate it.

When AI produces a few decisions, this is manageable.

Now imagine an enterprise where hundreds—or eventually thousands—of intelligent agents participate in workflows across IT, finance, procurement, customer service and supply chain.

If every meaningful action eventually reaches a human approval queue, the enterprise has not eliminated the coordination bottleneck.

It has simply moved it.

This may become one of the defining design challenges of the AI-native enterprise:

How do we increase autonomy without increasing uncontrolled risk?

Instead of depending on human review for every action, enterprises can increasingly encode authority, boundaries, controls, evidence and escalation into the environment in which AI operates.

Humans can then focus where judgment matters most: exceptions, ambiguity, high-consequence decisions and systemic patterns.

That is how trust begins to scale.

This Is a Leadership Question

Trust Architecture may sound like a technology problem.

It isn’t only one.

Technology teams can implement identity, permissions, observability and controls.

But they cannot independently determine how much authority an enterprise should delegate to AI.

That requires business judgment.

Business leaders must determine which decisions can be delegated.

Risk and compliance leaders must define acceptable boundaries.

Technology leaders must translate those boundaries into enforceable controls.

And executive leadership must establish accountability for the resulting system.

This leads to a question I believe more leadership teams will eventually confront:

Where are we willing to delegate authority to intelligent systems—and what must be true before we do?

That question moves AI governance beyond technology.

It places it directly into enterprise design.

A Question for the Boardroom

Imagine two enterprises.

Both have access to similar AI capabilities.

Both have Responsible AI policies.

Both have governance structures.

Both are experimenting with agentic systems.

But one still depends heavily on humans approving AI-generated actions.

The other has designed clear authority, decision boundaries, execution controls, observability and escalation into its operating environment.

Which organization will be able to scale AI-enabled execution faster?

More importantly:

Which will be able to scale it without losing control?

As AI capabilities become increasingly accessible, the differentiator may not simply be who has the most powerful models.

It may be who develops the organizational and technical infrastructure required to trust intelligent systems responsibly at scale.

Final Thought

The AI-native enterprise cannot depend on humans reviewing every intelligent action.

But neither can it depend on blind confidence in autonomous systems.

Something must exist between those extremes.

Designed trust.

Trust established through authority.

Strengthened by reliable context.

Bounded by decision rights.

Enforced through execution controls.

Made visible through observability.

Reinforced through human accountability.

As AI becomes part of enterprise execution, trust will no longer sit only around the technology.

Trust will become part of the infrastructure itself.

Continue the Series

This is Part III of Leading the AI-Native Enterprise.

Part I — The AI-Native Enterprise
How enterprise transformation changes when AI becomes part of the operating model.

Part II — The Enterprise Coordination Shift™
What changes when AI begins coordinating work rather than simply completing tasks.

Part III — Trust Architecture™
How enterprises can build confidence as intelligent systems increasingly participate in decisions and execution.

Next: Part IV — Human + AI Execution Models

What does enterprise execution look like when humans and AI increasingly work together?

Because the next challenge is not simply determining what AI can do.

It is designing how humans and intelligent systems should work together.

AI at the Frontlines explores how leadership, operating models, governance and enterprise design are evolving for the AI-native era.

~7-minute read

Editor’s Note: This article introduces The Enterprise Coordination Shift™, the first original framework in the Leading the AI-Native Enterprise series. It reflects my perspective on how AI is changing the economics of enterprise execution and why coordination may become the next enduring source of competitive advantage.

Every generation of enterprise leaders has been promised a technology that would fundamentally change how work gets done.

Scientific management standardized tasks.

Enterprise Resource Planning integrated business processes.

Cloud computing connected organizations.

Today, Generative AI promises another leap in productivity.

Across industries, organizations are investing in copilots, intelligent assistants, and automation platforms. Success is measured by familiar metrics: hours saved, tickets resolved, documents generated, or lines of code written.

These improvements are real.

But after observing enterprise AI transformations across industries, I believe we are asking the wrong question.

Instead of asking:

How can AI automate more work?

we should be asking:

How can AI help the enterprise coordinate better?

Because the biggest transformation AI will bring is not automation.

It is coordination.

The Productivity Trap

For decades, management has pursued a simple objective: make work more efficient.

That focus made perfect sense.

If every employee becomes more productive, the organization should perform better.

Yet most enterprise failures don’t occur because people are incapable of completing their work.

Projects slip because teams operate with different assumptions.

Critical decisions are delayed because the right information never reaches the right people at the right time.

Customer experiences suffer because every department optimizes locally while enterprise performance declines globally.

These aren’t productivity problems.

They are coordination problems.

AI unquestionably makes individual work faster.

But its greater opportunity lies in reducing the friction between people, teams, systems, decisions, and increasingly, intelligent agents.

That distinction changes how we should think about enterprise AI.

The Evolution of Enterprise Advantage

Every major era of management has optimized a different source of competitive advantage.

The Industrial Era focused on task optimization.

The Digital Enterprise focused on process optimization.

The AI-Native Enterprise must focus on enterprise coordination.

Task optimization improved individual performance.

Process optimization improved operational efficiency.

Enterprise coordination improves how knowledge, decisions, people, systems, and intelligent agents continuously align to deliver business outcomes.

This is not simply another technology upgrade.

It is a new management challenge.

Figure 1 — The Evolution of Enterprise Advantage™

Why Coordination Is Becoming Strategic

Modern enterprises have never had more information.

They have also never faced greater complexity.

Today’s leaders operate across global teams, distributed workforces, cloud platforms, regulatory environments, partner ecosystems, and increasingly, AI systems capable of generating recommendations and initiating actions.

At the same time, business moves faster than ever.

Decisions cannot wait for weekly governance meetings.

Dependencies rarely remain inside a single department.

Transformation is no longer an annual initiative.

It has become the operating environment.

The challenge is no longer collecting information.

It is coordinating it.

The organizations that coordinate knowledge, decisions, and execution most effectively will increasingly outperform those that simply automate more work.

Introducing The Enterprise Coordination Shift™

This observation has led me to a broader management hypothesis.

I call it The Enterprise Coordination Shift™.

For more than a century, organizations created competitive advantage by improving how work was performed.

First came task optimization.

Then process optimization.

The next frontier is enterprise coordination.

Automation remains essential.

But automation alone does not create an AI-native enterprise.

Coordination does.

The organizations that define the next decade may not be those deploying the greatest number of AI solutions.

They will be the ones coordinating people, knowledge, decisions, and intelligent systems more effectively than everyone else.

One Story Every Executive Recognizes

Imagine a global product launch.

Marketing believes everything is on schedule.

Engineering is tracking unresolved technical dependencies.

Legal is reviewing regional compliance.

Finance has not finalized pricing.

Sales has already committed delivery dates to customers.

Operations is preparing production.

Every team is performing well.

Yet the launch slips by another month.

No individual function failed.

The organization simply failed to coordinate.

Now imagine the same enterprise operating with shared context.

AI continuously identifies dependencies before they become risks.

Leadership teams see the same operational picture instead of conflicting dashboards.

Teams adapt plans dynamically rather than waiting for weekly status meetings.

Nothing in this scenario replaces human judgment.

Instead, AI reduces the friction separating planning from execution.

That is where enterprise value is created.

What This Means for Leaders

As AI becomes embedded in enterprise operations, leadership itself begins to change.

The defining question is no longer:

Where can we automate?

It is becoming:

How do we design organizations that coordinate humans and AI effectively?

Future leaders will not simply deploy AI technologies.

They will design operating models where knowledge flows across organizational boundaries, decisions are made with shared context, and intelligent systems amplify—not replace—human judgment.

In the AI-native enterprise, coordination becomes a strategic capability.

A Question Worth Taking to the Boardroom

Imagine walking into your next executive leadership meeting.

Which discussion creates greater long-term value?

How many additional tasks can AI automate this year?

Or...

How should we redesign our enterprise so knowledge, decisions, and execution move with dramatically less friction?

The first question improves productivity.

The second has the potential to redefine competitive advantage.

I believe that is the conversation every leadership team will eventually have.

Final Thought

Every era of management has been defined by a different source of competitive advantage.

The Industrial Era optimized work.

The Digital Era optimized processes.

The AI-Native Era will optimize coordination.

The organizations that define the next decade won’t necessarily be those with the most sophisticated AI models.

They will be the ones that coordinate knowledge, decisions, people, and intelligent systems better than everyone else.

That may become the defining advantage of the AI-native enterprise.

Continue the Series

This article is Part II of my Leading the AI-Native Enterprise series.

In the next edition, we’ll explore another question every executive will face:

Part III — Trust Architecture™

How do organizations build trust when humans and AI increasingly make decisions together?

If today’s AI transformation is organizational, trust will become one of its most important operating capabilities.

I hope you’ll join me for the next conversation.

AI at the Frontlines explores how leadership, operating models, governance, and enterprise design are evolving for the AI-native era. If these topics resonate with you, consider subscribing and joining the conversation as we build a practical management playbook for the next generation of enterprise transformation.

The AI-Native Enterprise Isn’t About Replacing Humans -It’s About Redesigning Enterprise Execution

What if the biggest AI transformation isn’t technological?

What if it’s organizational?

Most enterprise AI discussions still revolve around a familiar set of themes.

• Copilots.

• Automation.

• Productivity.

These conversations are important. But they may miss the larger transformation now emerging.

Over the last several months, I’ve become increasingly convinced that the future of enterprise AI may depend less on replacing human work and more on redesigning how enterprises themselves execute.

And that distinction may prove to be more profound than many leaders realize.

AI Is Changing Coordination More Than Tasks

The first wave of AI focused on helping individuals perform work.

The next wave may focus on coordinating work itself.

Historically, enterprises optimized tasks.

Processes were designed, standardized, measured, and improved. Automation reduced manual effort. Copilots enhanced individual productivity.

AI introduces a different possibility.

Intelligent systems are increasingly capable of coordinating activities across teams, functions, and workflows. They can interpret context, recommend actions, orchestrate execution, and help manage dependencies that previously required significant human coordination.

Consider a major incident.

Historically, coordination occurs through managers, bridge calls, escalation paths, and functional teams.

An AI-native enterprise may increasingly use intelligent systems to identify impacted services, assemble the right stakeholders, recommend remediation actions, and orchestrate execution across teams.

The challenge is no longer simply automation.

It is coordination at enterprise scale.

The question is no longer simply:

How do we automate work?

Increasingly, leaders may need to ask:

How should humans and intelligent systems execute work together?

That shift moves the conversation beyond productivity and toward enterprise execution itself.

Trust Architecture May Become as Important as Technology Architecture

As AI systems become more capable, the challenge facing enterprises is no longer capability alone.

It is confidence.

Leaders must determine:

• Where should AI assist?

• Where should humans approve?

• Where can systems operate independently?

• How are decisions monitored, governed, and explained?

Technology alone cannot answer these questions.

They require governance, accountability, transparency, and trust.

In many ways, trust architecture may become as important as technology architecture.

The organizations that succeed may not be those with the most advanced models.

They may be the ones that establish the clearest decision boundaries and the strongest mechanisms for trust.

AI Transformation Is Becoming Organizational Transformation

Digital transformation optimized processes.

AI transformation may redefine execution.

And execution sits at the heart of the enterprise.

As AI becomes embedded within operations, organizations may need to rethink:

• Leadership models

• Decision-making structures

• Governance mechanisms

• Operating models

• Human-machine collaboration

The implications extend well beyond technology.

They touch the fundamental ways organizations coordinate, decide, and execute.

Which means the future of enterprise AI may be less about deploying tools and more about redesigning the enterprise itself.

A Framework for Enterprise AI Evolution

The progression may look something like this:

The shift is not simply from manual work to automated work.

It is from isolated productivity to coordinated enterprise execution.

Three Themes Stand Out

As I look ahead, three themes continue to emerge:

AI is changing coordination more than tasks.

Trust architecture may become as important as technology architecture.

AI transformation is becoming organizational transformation.

These ideas are still evolving.

But they increasingly appear central to understanding what an AI-native enterprise may become.

Looking Ahead

We’re still early in this transition.

Many of the operating principles of the AI-native enterprise have yet to be defined.

But one thing appears increasingly clear.

The future competitive advantage of enterprises may not come from who deploys AI first.

It may come from who redesigns execution first.

Perhaps the most important question facing leaders is no longer:

“How do we deploy AI?”

But rather:

“How should enterprises themselves evolve because of AI?”

That may ultimately prove to be the more profound transformation.

And it is a question I will continue exploring throughout Season 5 of AI at the Frontlines.

About Season 5: Leading the AI-Native Enterprise

The next phase of enterprise AI may not simply be technological.

It may be organizational.

Throughout Season 5, I will explore:

• Human + AI execution models

• Trust architecture and governance

• Decision-making in AI-native organizations

• Leadership in the age of intelligent systems

• The future operating model of the enterprise

If these questions resonate with you, consider subscribing to AI at the Frontlines.

Because the future of enterprise AI may not simply be about technology.

It may be about how organizations themselves evolve.

AI at the Frontlines

Helping Leaders Navigate AI-Native Transformation.

How do enterprises themselves evolve because of AI?

Twenty editions later, the focus of AI at the Frontlines has evolved from models, tools, and copilots toward a bigger question:

How do leadership, operating models, governance, and organizational structures evolve in the age of AI?

As intelligent systems become increasingly capable, the deeper transformation may not be technological.

It may be organizational.

This flagship series explores how enterprises adapt to AI-native environments, examining the evolution of leadership, decision-making, coordination, governance, and the future of human and intelligent system collaboration.

Rather than focusing on isolated technologies, these essays examine the larger shift underway inside organizations themselves.

Topics explored in this series include:

1. The Bigger AI Transformation May Be Organizational

2. Leading the AI-Native Enterprise

3. Why AI-Native Enterprises Need New Leadership Models

4. AI Changes Coordination More Than Tasks

5. Human-in-the-Loop Is Becoming Human-on-the-Loop

6. The Rise of AI-Native Operating Models

7. Governance Is Becoming Strategy

8. Designing Organizations for Human + AI Collaboration

9. Decision-Centric Enterprises

10. The Future of Middle Management

This season represents the next chapter of AI at the Frontlines.

Because the most important transformation may not be AI itself.

It may be how enterprises evolve because of AI.

What started as a newsletter became a thesis.

Twenty editions later, the journey is just beginning.

What comes after copilots?

As AI capabilities continue to mature, the conversation is beginning to move beyond assistants and copilots toward systems that can reason, coordinate, and act with increasing autonomy.

This season explores the emergence of agentic systems and examines how enterprises may need to rethink trust, orchestration, and governance at scale.

From ServiceNow and Now Assist to AI control towers and trust architectures, these articles explore the foundations required for intelligent systems to operate safely and effectively within the enterprise.

Articles included in this season:

1. ServiceNow + Agentic AI

2. Now Assist: Copilot or Autonomy?

3. Trust Architecture

4. Where Agentic AI Delivers Value

5. The AI Control Tower

Beyond copilots lies orchestration.

How do intelligent systems evolve?

As AI systems become increasingly capable, the focus begins to shift from individual models and tasks toward autonomy, trust, and decision-making.

This season explores the emergence of intelligent systems and examines how enterprises may need to rethink control, oversight, and the relationship between humans and machines.

From reinforcement learning and autonomous enterprises to decision-making and AI-native reinvention, these articles explore the foundations of intelligent systems in the enterprise.

Articles included in this season:

1. RLAIF

2. The Autonomous Enterprise

3. Executive AI Decisions

4. Control Placement

5. AI-Native Reinvention

Stay at the Frontlines

Subscribe for free to receive new research, frameworks, and executive briefings on the transition to the AI-native enterprise. No paid subscription required.

Why do promising AI initiatives struggle?

While AI capabilities continue to advance, many enterprise initiatives struggle to move beyond experimentation.

This season explores some of the practical and organizational challenges that often prevent promising programs from achieving scale.

From pilot fatigue and shadow AI to reinforcement learning and reward systems, these articles examine why successful AI transformation requires more than technology alone.

Articles included in this season:

1. Pilot Fatigue

2. Shadow AI at Work

3. When AI Fails, What’s Your Plan B?

4. Why RLHF Changed Everything

5. Reward Models

Progress often stalls before it scales.

How do enterprises adopt AI?

This season captures the early lessons of enterprise AI adoption, from technical program management and task force design to scaling AI initiatives across the organization.

Articles included in this season:

1. Rethinking Technical Program Management in the GenAI Era

2. Building a GenAI + Agentic AI Task Force from Scratch

3. Cracking the 95% Failure Rate — A TPM’s Take on the MIT GenAI Divide Report

4. From GenAI to Agentic AI: Top 5 Pitfalls in Enterprise Rollouts

5. From Pilot to Platform: Scaling AI Across the Enterprise

   ---

Every transformation begins with foundations.

Twenty editions later, the focus has evolved from AI tools and copilots toward a bigger question.

How do enterprises themselves change because of AI?

What started as a newsletter gradually became something else. A body of work exploring leadership, operating models, governance, and organizational design in the age of AI.

Along the way, several ideas changed my thinking.

The Bigger Transformation May Be Organizational

Initially, I was fascinated by models, copilots, and automation.

But over time, it became increasingly clear that the deeper transformation may not be technological.

It may be organizational.

As AI capabilities mature, enterprises may need to rethink leadership structures, execution models, governance, and how humans and intelligent systems collaborate.

AI Changes Coordination More Than Tasks

Much of the discussion around AI focuses on automating individual tasks.

But the larger impact may come from how work gets coordinated across people and intelligent systems.

The future may be defined less by isolated tools and more by orchestration.

Increasingly, the question may no longer be:

“How do we automate tasks?”

But rather:

“How do we coordinate humans and intelligent systems at scale?”

Governance Is Becoming a Strategic Capability

Governance in the AI era is not about control.

It is about creating the conditions for trust, speed, and responsible innovation.

The organizations that build governance as a capability may ultimately move faster than those that treat it as a constraint.

Trust, accountability, and transparency are becoming competitive advantages.

Human Judgment Still Matters

Increasing autonomy does not eliminate the importance of people.

In many situations, human judgment becomes even more valuable.

The challenge is not replacing humans, but designing effective collaboration between humans and intelligent systems.

The future is unlikely to be human versus AI.

It is far more likely to be human with AI.

What Comes Next

The next phase of AI at the Frontlines will focus less on tools and more on enterprises themselves.

How organizations evolve.

How leadership changes.

How operating models adapt.

How intelligent systems reshape the future of work.

And how governance enables responsible autonomy.

Because twenty editions later, I believe the bigger transformation may still be ahead of us.

What started as a newsletter became a thesis.

Twenty editions later, the journey is just beginning.

TWENTY EDITIONS LATER

The focus has evolved from AI tools and copilots toward a bigger question:

How do enterprises themselves change because of AI?

AI at the Frontlines explores that question through essays, frameworks, and field notes on leadership, operating models, governance, and organizational design.

Written from the perspective of a practitioner with two decades of enterprise transformation experience, AI at the Frontlines is organized as a body of work rather than a collection of isolated articles.

What You’ll Find Here

Essays

Perspectives on leadership, operating models, governance, and the future of work.

Frameworks

Practical models for building AI-native enterprises.

Field Notes

Observations and lessons from two decades of enterprise transformation.

Season Series

A body of work organized around major questions facing enterprises in the AI era.

Current Seasons

Season 1 — Foundations

Building the base for enterprise AI.

Season 2 — Why AI Programs Stall

Understanding the barriers to scale.

Season 3 — Intelligent Systems

From automation to autonomy.

Season 4 — Beyond Copilots

From copilots to agentic systems.

Season 5 — Leading the AI-Native Enterprise

Designing and leading AI-native organizations.

What started as a newsletter became a thesis.

Welcome to AI at the Frontlines.