The Leader Who Doesn’t Know They’re Feared

Posted on by

 

There is a specific kind of dangerous leader that nobody talks about in boardrooms.

Not the one who shouts. Not the one who publicly humiliates people or rules through naked aggression. That leader is visible. People name them. Organisations deal with them eventually.

The dangerous one is the leader who has created a culture of fear without ever raising their voice. The one whose team has learned, quietly and carefully, that honesty is a liability. The one who genuinely believes they are approachable, open, and trusted, while the people around them have become experts at managing what they say, how they say it, and what they keep entirely to themselves.

This leader does not know they are feared.

And that is exactly what makes them so costly.

 

The Gap Between Who You Think You Are and Who They Experience

Most leaders who create fear do not intend to. That matters, and it also does not matter at all.

Intent does not determine impact. Culture does not care about your intentions. Your team’s behaviour is shaped entirely by what they have learned happens when they take a risk with you, when they bring bad news, push back on a decision, admit a mistake, or say the thing nobody else is saying.

The research on this is striking. Organisational psychologist Tasha Eurich conducted a multi-year study involving thousands of participants across multiple studies and found that while 95% of people believe they are self-aware, only 10 to 15% actually meet the criteria for genuine self-awareness. The finding most relevant here is this: internal self-awareness, how we believe we come across, and external self-awareness, how we actually come across to others, are essentially uncorrelated. A leader can be deeply reflective and still have almost no accurate sense of how they are experienced by their team. This is not an arrogance problem. It is a structural blind spot.

If those moments of honesty have gone badly, even once, even subtly, people remember. They adjust. They start self-censoring before the words leave their mouth. They develop what looks like professionalism but is actually armour.

And from your side of the table, everything looks fine. People are engaged in meetings. Nobody is causing problems. The reports are clean. The team seems to be functioning.

What you are actually seeing is a highly guarded team performing composure.

 

The Behaviours That Build Feared Leaders

None of what follows requires malicious intent. Most of it comes from competence, pressure, or simply the unchecked habits of someone who has always been the fastest, most decisive person in the room. The mechanism is consistent: a behaviour that felt like leadership produced a reaction in others that went unnoticed. And that reaction, over time, became the culture.

You respond to bad news with visible frustration. You do not think it is a big deal. You move on quickly. But the person who delivered that news felt the temperature drop the moment the words left their mouth. They filed it away. Next time, they will soften it. The time after that, they will delay telling you. Eventually, you will be the last to know.

You solve problems before people finish explaining them. You are fast and you are good and your instinct is usually right. But what the other person experiences is that their perspective does not matter. That the conversation is a formality. That you have already decided. So they stop bringing you half-formed problems. They only come when they have the answer, which means you lose access to the problems at the stage when you could actually help.

You confuse directness with dismissal. You believe you give clear, honest feedback. And you do. But there is a difference between feedback that challenges someone’s thinking and feedback that makes them feel their thinking is worthless. The words can be almost identical. The delivery is everything.

You reward the people who agree with you. Not deliberately. But your energy lifts when someone validates your thinking. Your attention sharpens. And the people in that room are excellent at reading energy. They learn the correlation. Agreement gets warmth. Challenge gets friction. The lesson lands fast.

You are always the most certain person in the room. Certainty from a leader is reassuring up to a point. Beyond that point, it signals that doubt is unwelcome. And if doubt is unwelcome, so is the information that generates it. Your team starts protecting you from complexity, which means they start protecting you from reality.

You have never once said: I got that wrong. Or if you have, it was so rare that people remember it. When a leader never admits error, the message to the team is clear: mistakes are not safe here. And a team that cannot make mistakes safely cannot take risks, cannot innovate, and cannot tell you the truth when the truth involves something going wrong.

 

 

What a Guarded Team Looks Like

The tragedy is that guarded teams can look like high-performing teams from a distance.

They are efficient. They deliver. They do not create noise. Meetings run smoothly because nobody says anything that might create friction. Reports look clean because the real information has been edited out. Problems get solved at the level below you because bringing them upward feels more dangerous than managing them quietly.

What you are missing is everything below the surface.

The risk that has been on someone’s mind for three weeks but has not made the register because the person managing it does not want to be seen as struggling. The team member who is six weeks from walking out the door because they have felt invisible for months. The supplier relationship that is quietly deteriorating because your account manager is telling you it is fine rather than managing the conversation you would need to have.

The gap between the reported reality and the lived reality grows, week by week, in proportion to how safe people feel telling you the truth.

And at some point, the gap becomes a crisis. And everyone knew except you.

 

Why the Leaders Who Need This Most Will Not Recognise Themselves

This is the most difficult feature of the whole dynamic.

Research by Jack Zenger and Joseph Folkman of leadership development firm Zenger|Folkman, drawing on one of the largest 360-degree feedback databases in existence, consistently finds that senior leaders show the widest gap between how they rate themselves and how their subordinates rate them, particularly on dimensions like listening, approachability, and empathy. The people with the most influence over organisational culture are, on average, the least accurate judges of their own impact on it.

The behaviours that create fear are almost always the same behaviours that made someone successful. The pace. The decisiveness. The high standards. The intolerance for mediocrity. These are not bad qualities. They are, in many environments, exactly what drove results.

But at a certain level of leadership, those same qualities, unchecked and unexamined, become the thing that limits the people around you. And because they drove your success, you do not interrogate them. You double down on them.

And the people around you adapt. They become mirrors. They reflect back what you want to see. You walk into rooms and the room agrees with you. You raise concerns and the concerns get managed. You ask if everything is on track and the answer is always, essentially, yes.

You look successful. The organisation is slowly becoming fragile.

The leaders who most need to hear this will read it and think of someone else. That is not a criticism. It is the nature of the blind spot. You cannot see what the system around you has been carefully constructed to hide from you.

 

The Only Way Through

It starts with an uncomfortable question, not asked of yourself, but asked of someone who will tell you the truth if the environment is safe enough.

What is it like to work with me when things are going wrong?

Not in general. Specifically when things are going wrong. When the pressure is highest. When the news is bad. When someone has made a mistake. What happens in that room?

If you do not know the answer, if you genuinely cannot predict what your team would say, that is the information.

The standard prescription at this point is 360-degree feedback. It can help. But the research on its effectiveness is mixed, and the most common pattern is familiar: leaders receive the results, rationalise the parts that are uncomfortable, and make short-term adjustments that do not hold. The tool is not the problem. The willingness to sit with what it reveals, and to keep sitting with it, is.

Because the leaders who have done the work know exactly how that room feels. They have been told. They have asked enough times, and created enough safety, that people have told them. Not through a survey. Through the kind of environment where someone can walk into your office and say: I need to tell you something you are not going to enjoy hearing.

The ones who have not done the work are always slightly surprised when good people leave. Slightly confused when the programme that looked fine on paper collapsed in execution. Slightly unsure why the energy in the room feels managed rather than genuine.

They are not bad leaders. They are unexamined ones.

And in the long run, the cost is exactly the same.

Pre-Mortem: The Pentagon’s Autonomous Drones Reset

Posted on by

 

The Pentagon’s Replicator programme promised thousands of cheap autonomous drones in two years and delivered hundreds. The response has not been to wind it down. It has been to dissolve it, rebuild it as a new command inside Special Operations Command, and ask Congress for roughly 240 times the money. A programme that under-delivered on a lean, fast model is being re-attempted on a vast one, and the case for why the second structure succeeds where the first did not has not yet been made in public.

A pre-mortem asks the same five questions, every time, applied to a current programme before failure is possible rather than after. This is the third in the series. The first looked at vendor accountability in regulated finance. The second looked at clinical safety accountability in regulated healthcare. This one looks at execution accountability in defence procurement, the hardest delivery environment of them all. Different sector, similar structural shape: commitment moving faster than the architecture meant to hold it to account.

 

The Bet

The bet is that scale fixes what speed could not. Replicator was announced in August 2023 with a target of multiple thousands of all-domain attritable autonomous systems inside roughly two years, run by the Defense Innovation Unit on about a billion dollars across two fiscal years. It was deliberately lean, built to route around the traditional acquisition machine. By the deadline it had fielded hundreds. The reset, the Defense Autonomous Warfare Group, carries a 2027 budget request of about $54 billion, against roughly $226 million the year before. The technical bet is sound on its face: mass autonomy is where warfare is going, and the United States cannot afford to be slow to it. The harder bet, the one sitting under the headline number, is that money and a command structure fix what was an execution problem. Those are different things, and the launch treats them as one.

 

The Assumption

One belief is doing all the work: that Replicator’s shortfall was a problem of resourcing and structure, solvable with more of both. The documented failures point elsewhere. Systems were selected that proved unreliable, too expensive, or too slow to manufacture at the quantities needed. Some existed only as a concept when they were chosen. And the programme could not procure software able to orchestrate and command large, mixed swarms of different drones, which is the actual technical heart of autonomy at scale. None of those is a budget problem. A bigger budget buys more of the same systems and more of the same integration gap. If the diagnosis is wrong, the cure scales the disease.

 

The Sequence

Commitment came before the architecture, again. Replicator launched in August 2023. A second line of effort, focused on countering small drones, was added by a Secretary of Defense memo in September 2024. The original thousands-by-2025 deadline arrived with hundreds delivered. The programme was then consolidated into a joint interagency task force, dissolved, and rebuilt as the new autonomous-warfare group inside Special Operations Command, with the first acquisition under the new structure landing in January 2026, two counter-drone systems. Only in April 2026 did the Secretary tell the House Armed Services Committee that a sub-unified command for autonomous warfare was coming. The command meant to own this is still being stood up around a commitment already made. The funding tells the same story. Of that $54 billion, only about $1 billion is appropriated base money. The other $53 billion is a request, parked in a flexible five-year reconciliation pot that Congress has not yet passed. The headline number signals overwhelming commitment. In hard terms it is roughly a billion dollars in hand and fifty-three billion in hope. The intention is real. The money, for now, is one dollar in every fifty-four.

 

The Pager

Start with the credit, because it is real. The new group has a named director, Lt. Gen. Francis L. Donovan (USMC), with a clear command line and an appointment made by the Secretary himself. That is more named, senior accountability than most large defence programmes ever put on the public record, and it counts for something. The harder question is operational and specific. Standing policy requires appropriate levels of human judgement over the use of force. At swarm scale, with attritable systems acting at machine speed, who is the named individual accountable when one of them engages wrongly? The command line is clear. The accountability for the autonomous decision itself, at the scale this programme is built to reach, has not been framed in public. A command answers for a programme. It is a harder thing to say who answers for a single autonomous engagement when there are thousands of them in the air.

 

The Proof

The committed measures are input measures. Dollars requested, units contracted, the first systems bought. There is no public outcome measure for capability actually delivered, no cost per effective intercept, no fielded-and-working-at-scale figure with a date attached. This matters because the proof problem already bit once. Leadership called Replicator on track in 2024 and said it had made enormous strides in 2025, while the independent accounting found hundreds, not thousands. When the people who own the programme also own the definition of progress, optimism outruns delivery. Second-attempt scepticism is earned, not unfair. In eighteen months, the question of whether this worked will be answered by whoever holds the platform to define what delivered at scale means, and right now that platform is a budget request.

 

Verdict

This is a serious programme with serious people behind it. The strategic logic is correct, mass autonomy matters and slowness is its own risk. The accountability has a name and a rank, which is rare. The first systems have been bought and are heading to the field. None of that is in doubt.

What is unproven is whether a command and a budget can fix a problem that was about manufacturing maturity, software orchestration, and realistic system selection. A reorganisation addresses none of those by itself.

The action is concrete. Publish the outcome measure, not the input: a fielded-and-working-at-scale metric with a date, committed before the reconciliation money is spent, not after. Name the human accountable for autonomous engagement decisions at scale, not only the command that owns the programme. And diagnose the first shortfall in public before scaling, so the much larger second bet rests on a corrected understanding rather than a hope.

If the department publishes a delivered-at-scale outcome measure tied to a named owner, and solves the swarm-orchestration software problem it could not solve the first time, this becomes the programme that proves autonomous capability can be fielded at speed. Without both, it becomes the most expensive way yet found to relearn that money and reorganisation do not fix an execution problem.

The Most Dangerous Status Report Is the One Everyone Is Comfortable With

Posted on by

 

The governance is running. The reports are flowing. The steering committee met on time, every question got a confident answer, and the pack looked clean.

And the programme is in more trouble than anyone in that room is prepared to say.

This is not an unusual situation. It is not a sign of dysfunction or dishonesty. It is, in my experience, the most common information environment in large-scale programme delivery. The data supports that reading. Research by Milliken, Morrison and Hewlin, published in the Journal of Management Studies, found that 85% of employees have withheld important information from their manager because they feared the consequences of speaking up. The fear is not of formal punishment. It is relational: the fear of being seen negatively, of damaging a relationship, of being labelled someone who creates problems rather than solves them.

This is not a minority behaviour. It is the default.

The question is not whether a filter exists on your programme. The question is how thick it has become, and whether you would know.

 

Nobody Decides to Build the Filter

The pattern I have watched play out more times than I can count begins with a capable, experienced leader who genuinely means what they say. They have told the team, in kick-offs and town halls and one-to-ones, that they want to hear the bad news early. They are not performing openness. They believe it.

But then someone raises a concern in a steering committee and the leader’s body language shifts before the words are out. A risk gets flagged and the first question is why it was not caught earlier rather than what needs to happen now. A project manager delivers a difficult update and spends the following week under a level of scrutiny that has nothing to do with fixing the problem.

Nobody announces a new policy. Nobody says: do not bring me bad news.

But the room notices. Every single time.

And slowly, without anyone deciding to do it, the filter gets built. The team learns which concerns land well and which ones create friction. They learn how to frame things to reduce the emotional temperature in the room. They learn the difference between the truth and the version of the truth that keeps the meeting moving and their professional standing intact.

The updates keep arriving. The reports keep flowing. The governance keeps running.

But the signal has been stripped out. What remains is noise dressed up as information.

 

You Do Not Build This Through Negligence

This is the part that most leadership development will not tell you directly. You do not build a closed information environment through negligence. You build it through a series of entirely human, entirely understandable responses to difficult moments.

A flash of impatience when a problem arrived at the wrong time. A habit of moving to solutions before the problem is fully understood. A preference, however subtle, for the reassuring narrative over the complicated one.

These are not character flaws. They are instincts under pressure.

But at leadership level they are not private. The CIPD’s 2024 evidence review on psychological safety identifies leader and manager behaviour as the most critical driver of whether people feel safe to speak up, and specifically notes that what matters is not what leaders say about wanting honesty, but what they demonstrate through their actions when honesty arrives. The research is unambiguous: psychological safety is fragile. A single punitive response to good-faith feedback can damage trust that took months to build.

Every reaction is observed, interpreted, and factored into how safe it feels to tell you the truth next time. The leader who says they want honesty but visibly struggles to receive it is not running an open culture. They are running an organisation that has learned to give them what they can handle rather than what they need.

The most dangerous status report is not the one with red items on it.

It is the one everyone is comfortable with.

 

Four Practical Moves

The filter is not permanent. It is a learned behaviour, and learned behaviours can be unlearned. But reversing it requires something more specific than an open-door policy.

Stop asking questions that invite the managed answer. “How are things going?” will get you the curated version every time. Try instead: what is the one thing you would not put in a status report but think I should know? If this programme were going to fail, what would the early sign look like? What are we not talking about that we probably should be? Those questions signal that you are interested in the reality, not the performance of it.

Go to where the real work is happening. Not to inspect. To listen. The people closest to delivery carry an understanding of programme health that rarely makes it into formal reporting. A single honest conversation with a delivery lead or a technical team that has been carrying a quiet problem for weeks will tell you more than three months of steering committee updates.

Create a visible moment where surfacing difficulty is rewarded rather than merely tolerated. When someone raises something uncomfortable and your public response is genuine appreciation followed by a real conversation about what to do next, the entire room recalibrates what is safe to say. One moment like that shifts the culture more than any open-door policy ever will. The inverse is equally true: one moment where the messenger suffers sets the filter back months.

Learn to read silence as data. The steering committee where every question gets a confident answer. The risk log that has not changed in three weeks. The team that delivers polished updates but never raises anything unexpected. These things can mean a programme is running well. They can also mean the filter is fully operational and the real conversation is happening somewhere else entirely. If nobody is telling you anything that surprises you, that is not necessarily a sign that everything is on track. It may be a sign that you have stopped being the kind of leader people bring hard news to.

 

The One Question That Cuts Through

There is a question I now use when a programme looks clean but feels wrong.

I find someone close to the real work. Someone who has been there long enough to know where the bodies are buried. And I ask them one thing.

What does everyone here know that nobody is saying out loud?

The answer to that question is almost always where the programme actually is. The gap between that answer and what appears in the formal reporting is almost always where the real leadership work needs to happen.

 

Comfortable Information Is Borrowed Time

PMI’s research on complex programme delivery is consistent on this point: early warning signals are frequently present and frequently ignored, causing problems to compound in severity before they are addressed. The pattern is not exceptional. It is systematic.

Every week a real problem stays hidden is a week where the options for addressing it narrow. Manageable risks become serious ones. Recoverable situations become critical ones. And the longer the filter operates, the more the team’s trust erodes, because people who know the truth and watch it go unacknowledged eventually stop believing that leadership is operating in good faith.

When the programme finally tells you the truth, and it always does eventually, the question is rarely how to get back on track.

It is whether getting back on track is still possible.

 

The Leaders Who Get This Right

The leaders who consistently deliver in high-stakes environments are not always the most experienced or the most technically skilled.

But they share something that is harder to develop than either of those things. They have learned to want the truth more than they want to be comfortable. They have built the self-awareness to notice when they are receiving a managed version of reality, and the discipline to go looking for the unmanaged one. They have created environments where people bring problems early because they have learned, through consistent experience, that doing so leads somewhere useful.

That is not a natural state for most leaders. It requires sustained effort, genuine self-awareness, and a willingness to sit with difficult information and resist every instinct to make it someone else’s problem.

But the alternative is a programme that looks healthy until it does not. A team that has learned to give you what you can handle. A steering committee that runs on time and misses everything that matters.

When your team tells you how things are going, are they telling you what is happening?

Or are they telling you what they have learned you can live with?

The gap between those two answers is where most programmes are won or lost.

Why Western Delivery Frameworks Stall in the Middle East (and What to Do Instead)

Posted on by

 

I remember sitting across from a senior government official in the Gulf, about six weeks into a major transformation programme. On paper, everything was moving. The governance framework was in place. The workstream leads had been assigned. The project plan had been reviewed and signed off. The first steering committee had gone smoothly.

And yet nothing was actually happening.

Not because of incompetence. Not because of a lack of resources. Not because the methodology was flawed. The team I was working with was experienced, capable, and well-intentioned. But they had arrived with a delivery model built for a different context, and they were applying it with the confidence of people who had never had reason to question it.

That pattern is playing out at an extraordinary scale right now. Saudi Arabia’s ICT market surpassed $48 billion in 2024, the largest technology market in the Middle East. McKinsey’s 2025 State of AI in GCC Countries report, drawing on surveys of senior GCC executives, found that 84% of GCC organisations have adopted AI in at least one business function, and only 31% have successfully scaled or fully deployed across the organisation. That is a 53-percentage-point gap between starting and delivering, across some of the best-funded, most ambitious transformation programmes in the world.

The question is not why organisations in the region are investing. The question is why so much of that investment stalls between intention and outcome.

After more than a decade delivering in the Middle East, I think I know the answer. And it is not the one most people reach for.

 

The Assumption That Travels Badly

Western delivery frameworks, whether PRINCE2, PMI, SAFe, or the various proprietary methodologies that large consulting firms carry from engagement to engagement, are not neutral tools. They are cultural artefacts. They were built in specific organisational contexts, shaped by particular assumptions about how decisions get made, how accountability flows, how disagreement is handled, and what progress looks like.

Those assumptions are rarely stated explicitly. They do not need to be, in the environments where these frameworks were designed. Everyone in the room already shares them. But the moment you move those frameworks into a fundamentally different cultural context, the unstated assumptions become the problem.

Hofstede’s cultural dimensions research (now maintained by The Culture Factor Group) offers a useful lens here. Arab countries consistently score high on two dimensions that bear directly on programme delivery. The first is power distance: the degree to which authority is respected rather than openly challenged, and the degree to which the most senior voice shapes the room rather than the most technically accurate one. The second is uncertainty avoidance: a preference for predictability, a resistance to ambiguity, and a reluctance to surface risk that might destabilise a process already formally endorsed. These are not character flaws or cultural limitations. They are consistent patterns that predict specific delivery behaviours imported frameworks are simply not designed to manage.

The typical Western delivery model assumes a relatively flat decision-making structure where the person with the most relevant expertise speaks most loudly. It assumes that challenge and disagreement in a meeting are signs of healthy engagement rather than disrespect. It assumes that formal sign-off is the meaningful moment of commitment and that what is agreed in the room will be actioned after it. It assumes that timelines create accountability and that accountability creates action.

In the Middle East, several of those assumptions do not hold. And the teams that arrive without understanding this do not fail because they lack capability. They fail because they are solving the wrong problem.

 

What Actually Drives Delivery in This Region

Execution in this region is not primarily operational. It is relational.

This is not a cultural curiosity or a soft consideration to be acknowledged in a pre-departure briefing and then set aside. It is a delivery requirement. Understanding it, genuinely understanding it rather than paying lip service to it, is the difference between a programme that moves and one that generates activity without progress.

Decisions in many Middle Eastern organisations, particularly in government and quasi-government entities which dominate the regional landscape, do not flow through the formal governance structure in the way a Western framework assumes. The steering committee may ratify decisions, but the real alignment happens elsewhere. In relationships that have been built over time, in conversations that take place outside the formal meeting structure, in the space between hierarchy and trust that no project plan captures.

Hierarchy here is not an obstacle to navigate around. It is the delivery infrastructure. Understanding who the real decision-makers are, what they care about, how they receive information, and what kind of relationship needs to exist before they will move is not supplementary to the delivery approach. It is the delivery approach.

Equally, the pace at which genuine commitment forms is different. A Western programme manager reads a signed-off plan as a committed baseline. In many regional contexts, that same sign-off is closer to the beginning of a conversation than the end of one. Real commitment, the kind that produces action, is built through repeated engagement, demonstrated respect, and a track record of following through. It cannot be manufactured by a governance process, however well designed.

 

The Meeting That Agrees to Everything and Changes Nothing

One of the most consistent patterns I encounter when stepping into stalled programmes in the region is what I have come to think of as performative alignment. The meetings happen. The presentations are well received. The heads nod. The action items are recorded. And then the follow-through does not come, not because anyone has decided not to cooperate, but because the alignment that appeared to exist in the room was not the deep kind that produces action.

In high-context cultures, of which many in the Middle East are clear examples, direct disagreement in a formal meeting setting carries a social cost that most Western delivery professionals underestimate. Saying no to a proposal in front of a room of peers and seniors is not simply a professional difference of opinion. It can feel like a breach of the respect and harmony that the meeting is partly there to maintain.

The result is that concerns, reservations, and genuine blockers often do not surface in formal governance forums. They emerge later, in quieter conversations, or they do not emerge at all. The experienced regional delivery leader learns to read what is not being said in a meeting as carefully as what is. The silence after a proposal is not always agreement. The smooth meeting is not always a sign of progress.

Teams that do not understand this dynamic spend months wondering why a programme that looked aligned keeps stalling. The answer is usually that the alignment they measured was the visible kind, and what drives delivery here is the invisible kind.

 

Where International Teams Get It Wrong

The failure mode I see most consistently is not incompetence or arrogance, although both exist. It is the application of a known model to an unknown context, with too much confidence and too little curiosity.

The international team arrives. They bring the framework, the templates, the governance structure, the reporting cadence. They run the kickoff. They establish the workstreams. They hold the first set of meetings. Everything looks like it is moving. The client counterparts are polite, engaged, and apparently aligned.

Then the programme slows. Decisions that should take days take weeks. Approvals that seemed close keep getting deferred. Stakeholders who appeared committed become harder to reach. The team escalates. They add more governance. More reporting. More pressure. The programme slows further.

What they are experiencing is the consequence of building delivery infrastructure without first building relational foundations. They have a governance model without trust underneath it, and governance without trust is just paperwork.

The other common failure is treating local counterparts as recipients of the methodology rather than as partners in understanding the context. The best people in any regional organisation carry an understanding of how things actually work, who the real influencers are, where the genuine blockers sit, and what has been tried before and why it did not land. Engaging that knowledge seriously, rather than as a courtesy, would transform the delivery approach. Most international teams access about ten percent of it.

 

What to Do Instead

The answer is not to abandon rigour or to conclude that structured delivery does not work in the region. It does. But it works differently, and the sequence matters enormously.

The first investment, before the governance framework, before the project plan, before the first steering committee, is in relationships. Not networking in the transactional sense. Genuine relationship-building, rooted in curiosity and respect, that creates the conditions under which honest conversation becomes possible. In a region where trust precedes transaction rather than following it, the time spent on this is not a delay to delivery. It is the foundation of it.

The second shift is in how decisions are understood and pursued. Rather than designing a governance structure and expecting decisions to flow through it, the experienced regional programme leader maps the real decision-making landscape. Who are the individuals whose genuine endorsement will move things? What do they need to see, hear, or feel before that endorsement becomes real? What informal conversations need to happen before the formal ones? Answering those questions honestly, and building a relationship strategy around them, is more valuable than any governance framework.

The third adjustment is in how alignment is tested. Rather than reading smooth meetings and nodding heads as confirmation of commitment, effective regional delivery leaders build in deliberate mechanisms for surfacing the real picture. Private conversations with key counterparts after formal sessions. Trusted intermediaries who can carry honest feedback in both directions. An explicit understanding that the formal meeting is often where positions are displayed rather than where they are formed.

Fourth, the pace of delivery needs to be calibrated to the pace at which genuine alignment forms, not the pace that the project plan demands. This is uncomfortable for Western programme managers trained to treat a timeline as a commitment. But the cost of false pace, the appearance of movement without the substance of it, is far higher than the cost of taking the time to build the real thing.

Finally, and perhaps most importantly, local knowledge must be treated as a strategic asset rather than a logistical courtesy. The people who understand the organisation, the culture, the history of what has been attempted before, and the unwritten rules that govern how things actually get done are the most valuable resource on the programme. Structuring the delivery approach around that knowledge, rather than around the imported framework, is the shift that most changes outcomes.

 

The Deeper Lesson

Delivering in the Middle East has taught me something that has made me a better programme leader everywhere, not just in the region.

Context is not a complicating factor. It is the medium through which all delivery happens. Every organisation has its own version of the unwritten rules, the informal power structures, the historical sensitivities, and the cultural patterns that shape how work actually gets done. The Middle East makes these visible in ways that Western environments sometimes obscure, because the gap between the imported model and the local reality is large enough that you cannot ignore it.

But the principle is universal. The best delivery professionals I have worked with anywhere in the world are the ones who arrive with curiosity before they arrive with answers. Who treat understanding the environment as the first act of delivery, not a preliminary to it. Who know that a framework is a starting point, not a solution.

The frameworks that travel well are not the ones with the most sophisticated methodology. They are the ones held lightly enough to be adapted, by people self-aware enough to know when adaptation is what the moment requires.

 

A Final Thought

If you are leading a programme in the Middle East and it has the shape of movement without the substance of it, the instinct will be to add more structure, more governance, more reporting, more pressure. That instinct is usually wrong.

The question to ask instead is simpler and harder. Do the people who need to move this programme forward trust you enough to tell you what is actually in the way? Have you built the relationships that make honest conversation possible? Do you understand not just the governance landscape but the human one?

If the answer to any of those questions is uncertain, that is where the work is. Not in the framework. Not in the plan.

 

Pre-Mortem: NHS Frontline Productivity Programme

Posted on by

 

On 1 April 2026, NHS England formally launched the Frontline Productivity Programme. It succeeds the £2 billion Frontline Digitisation Programme and is anchored to the NHS 10-Year Health Plan. The headline target is a 2% year-on-year productivity gain over three years. The lead use case is Ambient Voice Technology (AVT), AI-powered ambient scribing for clinicians, with £200 million committed in year one. The Department of Health and Social Care (DHSC) and NHS England have appointed Rob Thompson as joint Chief Digital, Data and Technology Officer.

A pre-mortem asks the same five questions, every time, applied to a current programme. This is the second in the series. The first looked at vendor accountability in regulated finance. This one looks at clinical safety accountability in regulated healthcare. Different sector, similar structural shape.

 

The Bet

The NHS is betting that AVT can deliver enough of the 2% year-on-year productivity gain to justify scaling deployment to tens of thousands of clinicians faster than the clinical safety framework for AI-enabled ambient scribing can be ratified. The technical bet rides on multi-site evidence led by Great Ormond Street Hospital (GOSH) across nine London NHS sites and 17,000 patient encounters: a 23.5% increase in patient interaction time, an 8.2% reduction in appointment length, and a 13.4% increase in A&E patients per shift. The strategic bet is that 19 self-certified suppliers competing for trust contracts will produce price discipline without producing safety variance. Reasoned bets, made under genuine pressure, backed by measurable evidence. But they are bets, and the framing reads as inevitability.

 

The Assumption

One belief is doing all the work: that clinicians using AVT will verify AI-generated notes against the patient context every time, at scale, rather than develop the same review-as-rubber-stamp pattern automation has produced in every regulated environment it has reached. The mechanism that produces the productivity gain is the same mechanism that erodes clinical attention to the note. If review thins because AVT proves “good enough” most of the time, the productivity number stays positive while clinical safety quietly degrades. Patient Safety Learning argued earlier this year that Copilot has arrived in the NHS without the operational guidance clinicians need to use it safely.

 

The Sequence

Capability shipped before the operational governance for AI-enabled ambient scribing was ratified. South West London is rolling out AVT to 20,000 clinicians across four trusts. University Hospitals of Leicester and Northamptonshire have deployed to over 10,000. Hertfordshire Community NHS Trust has moved past pilot to full rollout. NHS England published a 19-supplier self-certified AVT registry in January. Underneath, the clinical safety standards DCB0129 and DCB0160 are under active review, and the Explainability-Enabled Clinical Safety Framework for AI is still being developed. Commitment came first. The assurance framework is catching up.

 

The Pager

The accountability layer on this programme is more developed than most national digital programmes ever achieve. Rob Thompson holds a joint DHSC/NHSE Chief Digital, Data and Technology Officer post: senior, named, public, accountable. Chief Clinical Information Officers (CCIOs) at every deploying trust carry statutory DCB0160 deployment accountability. That deserves credit. The harder question is operational. When an AVT-generated note contains a clinically significant error that affects patient care, who is the named individual who carries the pager that night? The trust CCIO? The supplier on the registry? The clinician who signed off the note? The accountability is statutory; the operational reporting line for AI-specific clinical safety failure has not yet been publicly framed for AVT.

 

The Proof

Three outcome measures sit in the public record: the 2% year-on-year productivity gain, the GOSH-led multi-site evaluation, and the Oxford University Hospitals pilot in which 90% of clinicians reported reduced documentation time. All three measure clinician time and patient throughput. None measure clinical safety. A 2025 national cross-sectional study in the Journal of Medical Internet Research (JMIR), covering 178 NHS organisations and 14,747 digital health technology deployments, found that only 17.3% were fully assured against both DCB0129 and DCB0160. At a typical NHS trust, only 24.5% of deployed technologies held both assurances. The standards exist. Compliance with them is patchy. There is no committed measure for AVT-attributable adverse event rate by supplier, the rate at which clinicians materially amend AI-generated notes versus accept them, or DCB0160 compliance inside the AVT registry specifically. In 18 months, “did this work?” will be answered by whoever owns the platform to define what safe enough means.

 

Verdict

The Frontline Productivity Programme is more carefully constructed than most NHS technology programmes of the past two decades. Named senior accountability, real pilot evidence, multiple trusts in genuine production deployment, a clear use case the workforce wants. None of that is in dispute.

What is in dispute is whether the underlying clinical safety assurance layer holds at scale. DCB0129 and DCB0160 exist. Compliance with them currently runs at a quarter of what it should be. The deployments are racing toward 20,000-clinician scale while the AI-specific framework is still being written.

The action is concrete. Name the human at each deploying trust who carries the pager when an AVT-generated note causes patient harm. Demand per-supplier clinical safety performance reports from each of the 19 registry vendors, not self-certifications. Publish a clinical safety outcome measure alongside the productivity target before the year is out: adverse event rate change attributable to AVT, broken out by trust and by supplier.

If NHS England publishes a clinical safety outcome measure tied to a named owner in six months, and the AVT registry shifts from self-certification to audited compliance, the Frontline Productivity Programme becomes a model for AI deployment in regulated public services. Without both, the productivity number stays positive while the question of whether it was worth the clinical safety risk remains structurally unanswerable.

Your Personality Type Is Not Your Leadership Style. Your Behaviour Under Pressure Is.

Posted on by

Every few months, a personality framework resurfaces in leadership circles. DISC. Myers-Briggs. Enneagram. Belbin. Someone shares an assessment, teams complete a quick questionnaire, and for a week or two there is a flurry of conversation about whether the programme director is a high-D or the CIO is an INTJ. Then the work continues, and the framework quietly fades until the next cycle begins.

The problem is not that these frameworks are useless. Some of them are genuinely valuable as self-reflection tools and conversation starters. The problem is the assumption baked into almost every one of them: that your personality type shapes your leadership style in a consistent, predictable way. That if you are a Dominance type, you lead like Elon Musk. That if you are a Steadiness type, you lead like Satya Nadella.

The research says otherwise. And so does twenty years of watching this play out on real programmes, in real organisations, under real pressure.

 

The Landscape Is Larger Than the Label

DISC is one of a dozen personality-to-leadership models in active commercial use. MBTI classifies people across four dimensions derived from Jungian typology and is used by the vast majority of Fortune 100 companies, despite repeated criticism of its scientific validity. The Big Five model (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) has the strongest empirical backing of any personality framework but is the least commercially dominant, partly because traits measured on a continuous spectrum are harder to communicate at a glance. Belbin maps nine team roles, not personality types, which is a meaningful distinction. CliftonStrengths focuses on amplifying what you already do well rather than categorising who you are.

Each has a legitimate use. The problem begins the moment organisations treat them as performance predictors rather than self-awareness tools.

A landmark meta-analysis by Judge and colleagues in 2002, examining more than seventy studies across eleven thousand leaders, found that Conscientiousness and Extraversion were the strongest personality predictors of leadership performance, Conscientiousness most consistently linked to sustained effectiveness, Extraversion most linked to leadership emergence. Not type. Not style. Traits: the disposition to follow through, to maintain standards, to engage and bring others along. That finding has held up across replication after replication since.

 

The Attribution Problem

Personality frameworks routinely attach well-known leaders to their quadrants. The Dominance leader drives results and moves fast, so Elon Musk becomes the example. The Influence leader inspires and builds momentum, so Oprah Winfrey is named. The Steadiness leader creates stability and builds trust, so Satya Nadella is cited. The Conscientiousness leader ensures accuracy and high standards, so Mark Zuckerberg is assigned.

These attributions are speculative. As far as I know none of those individuals has publicly shared a verified assessment result. The assignments are made by inference from visible behaviour, which is circular reasoning: the framework predicts the behaviour, the behaviour confirms the framework, and nothing has actually been tested.

More importantly, the leaders themselves tend not to describe their effectiveness in terms of type. Satya Nadella, in Hit Refresh, describes empathy not as a natural trait but as a capability he had to actively develop, transformed in part by his experience as a father to a son with cerebral palsy. He credits that deepening, not his personality type, as the foundation of the cultural shift at Microsoft. The distinction between those two things, personality and the deliberate cultivation of capability, is the difference between a leader who peaked and a leader who grew.

 

Adaptability Is the Variable That Actually Predicts Success

Research on Emotional Intelligence shows that the ability to read a situation and modulate your response is among the strongest predictors of transformational leadership success, more practically relevant in complex change environments than a fixed personality classification.

The failure patterns make this concrete.

The high-D leader who drove results at pace in the first two quarters, and then, when the programme hit a wall and the team needed to be heard, doubled down on speed and ownership rather than shifting to inclusion and trust-building. The team read it as pressure, not leadership. Momentum collapsed.

The high-I leader who was exceptional at generating excitement and alignment across stakeholder groups, but whose follow-through was inconsistent. Commitments made in workshops were not tracked. The programme office spent six months chasing actions that the leader had already moved on from.

The high-S leader, calm and supportive under normal conditions, who then avoided a critical conversation with a failing vendor for three months because confrontation felt incompatible with their style. By the time the conversation happened, the programme had lost time it could not recover.

The high-C leader whose risk documentation was genuinely thorough, dependencies mapped, decision papers written to a standard the programme could stand behind, and who then spent ten weeks refining a business case while the implementation window closed around them. Not because the analysis was wrong. Because the level of certainty required before acting was higher than any live programme can ever provide, and by the time the final version was approved, the decision had already been made by events.

These are not personality failures. They are adaptability failures.

 

The Section Worth Building a Leadership Practice Around

Every serious personality framework identifies failure modes for each type. Not how the type shows up at its best, which tends to be flattering and self-affirming, but how it creates risk for the people around it and the outcomes it is responsible for when left unchecked.

For Dominance: moving too quickly without bringing others along. For Influence: energy is high but follow-through can vary. For Steadiness: harmony can be prioritised over progress. For Conscientiousness: the pursuit of accuracy can delay decision-making past the point where decisions still matter.

Every one of those is a real, recurring failure mode in transformation environments. The practical question is not which type you are. It is how aware you are of your natural failure mode, how quickly you can recognise when you are heading towards it, and whether you have built the discipline to shift before the damage is done.

 

What to Do with a Framework Once You Have One

Use it as a mirror, not a map. A mirror shows you how you naturally show up. A map tells you where to go. The frameworks are good mirrors. They are poor maps.

The leaders I have seen sustain high performance through complex, multi-year programmes were not distinguished by their type. They were distinguished by self-awareness, by a willingness to receive genuine feedback, and by the disciplined habit of asking, particularly under pressure, whether their natural response was the right one for the situation.

Knowing you are a Dominance type is not a leadership strategy. Knowing that your instinct under pressure is to accelerate when the situation requires you to slow down, and having the discipline to catch yourself before it costs you. That is.

AI Is Eating Theory. Companies Are Firing the People With Judgement.

Posted on by

 

AI is replacing the work that used to define the first decade of a career. At the same moment, organisations are quietly thinning out the people whose work AI cannot do.

That pairing sits somewhere between obvious and uncomfortable, depending on which part of the workforce you sit in. The data behind it is no longer disputed. The talent decision being made in response to it is almost universally backwards.

 

What AI is actually replacing

The popular framing of AI in the workplace is that it threatens knowledge workers broadly. The 2025 data tells a sharper story.

Stanford’s Digital Economy Lab released a study in November 2025 with the deliberately unsettling title Canaries in the Coal Mine. It tracked employment in occupations highly exposed to AI from late 2022 onwards. Workers aged 22 to 25 in those roles saw employment fall by roughly 13%. Workers in their forties, fifties and sixties in the same occupations continued to grow.

The mechanism is not redundancy. It is non-replacement. Entry-level vacancies are quietly not being backfilled. The career ladder is losing its bottom rungs.

The Stanford authors are unusually direct about why. AI is replacing codified knowledge, the part of expertise that can be written down, while complementing the experiential wisdom that only comes from years on the job. Other 2025 work in customer support and software development tells the same story. AI lifts the bottom of the distribution faster than the top. Two-month-experience workers using AI now match six-month-experience workers without it. The work AI does best is the kind of standardised, learn-from-a-book task that used to define the first few rungs of a career.

 

The thing AI cannot replicate

There is a second half to this story that gets less coverage.

Boston Consulting Group ran a study with Harvard Business School using 758 of its own consultants and GPT-4. On standard tasks, AI users completed 12% more work, 25% faster, with 40% better quality. The finding that rarely makes the press summary: when the same study tested tasks designed to fall outside the model’s actual capability, consultants using AI were 19 percentage points less likely to produce correct solutions. AI made experts wrong more often when the problem required judgement AI lacked.

The capability to know when to trust an AI answer and when to override it is itself a function of experience. It is built from a personal library of cases, situations and outcomes that no model has been trained on.

Decades of research in naturalistic decision-making, the field Gary Klein founded by watching firefighters and military commanders make calls under uncertainty, describes the same mechanism. Experts under pressure do not deliberate between options. They pattern-match against situations they have seen before. The library is built by exposure, not by reading frameworks.

This is what is meant by judgement. It is the residual human advantage in the AI era, and it has a clear demographic profile.

 

The talent decision being made backwards

Put the two findings beside each other.

AI is removing the codified, junior-level work fastest. The cohort whose work AI is actually complementing is the experienced one. The economic logic of an organisation in 2026 should be to lean into that experienced layer, because it is the part of the workforce AI cannot reproduce and which increasingly determines the quality of any AI-augmented output.

What organisations are doing instead is the exact opposite. The over-50 cohort is being quietly thinned through restructures, voluntary exit programmes, redundancy schemes, and the slow erosion of roles experienced workers tend to hold. It is rarely a stated policy. It is almost everywhere a pattern.

The talent decision is being made backwards. The cohort being pushed out is the one most worth keeping. The cohort being squeezed at the bottom is the one whose work AI is already doing. The organisation ends up with no future and no memory.

 

The cost of forgetting

There is an institutional dimension to this that gets ignored because it does not show up in the next quarterly report.

Roughly 42% of an organisation’s working knowledge sits in the heads of individual employees and nowhere else. Industry estimates put the cost of knowledge loss from rapid organisational change at tens of billions of dollars a year across Fortune 500 firms. The direction is consistent even where the precise figure varies. Restructures remove people, and the people take the unwritten knowledge with them. A newly arrived CEO who clears out the over-50 cohort does not just lose those individuals. They lose the only group who remembers why the last three transformations failed and what is different about this one.

That is not a fairness argument. It is a structural one. The organisation is paying a real cost. It will appear on the books eighteen months later, in the form of mistakes the experienced layer would have caught.

 

What we are not calling it

Age discrimination is unlawful in most major jurisdictions. It is also one of the most reliably under-reported categories of workplace harm, because it is rarely framed as discrimination by the people doing it.

ProPublica’s multi-year investigation into IBM found the company eliminated more than 20,000 workers aged 40 and over from 2013 onwards. The US Equal Employment Opportunity Commission concluded in 2020 that the layoffs had a clear adverse impact on older workers. More than 85% of those targeted for layoff in that period were older workers, even when rated as high performers. In 2023, former IBM HR professionals filed suit alleging termination linked to age and explicit plans to replace them with AI.

Almost no one running these processes describes them as age discrimination. They are called “right-sizing,” “talent refresh,” “succession planning,” “rebalancing the pyramid.” The language and the outcome have been routinely diverging for at least a decade. What is new in 2025 is that AI has made the underlying decision economically illiterate as well as legally exposed.

 

Where this leaves you

If you run an organisation that has quietly pushed out the experienced cohort, you have spent real money to remove the layer of your workforce AI cannot replicate, while leaving in place the layer whose work AI is doing without you noticing.

The question is not whether you can afford to keep experienced staff. It is whether you can afford to lose them at exactly the moment they became the most valuable people on your payroll.

Healthy Organisations Are Built by Healthy Leaders

Posted on by

 

Healthy organisations are not the product of wellbeing programmes. They are the product of leadership behaviour. Six disciplines describe what that behaviour actually looks like.

The corporate wellbeing market has more than doubled in the last decade. Employee wellbeing scores have flatlined. Burnout, disengagement, and quiet quitting continue to climb in almost every survey that measures them. None of this is for lack of effort. It is for lack of diagnosis. The budget has gone to the wrong department.

 

Wellbeing is downstream of how leaders behave

When a workplace is unhealthy, the temptation is to add a programme. Meditation apps. Resilience workshops. Mental health days. Employee assistance schemes with a freephone number nobody calls. These have their place, but none of them touch the actual cause.

What makes a workplace unhealthy is almost always the operating environment leaders create. The expectations they signal. The behaviours they reward. The hours they keep. The candour they tolerate. The boundaries they ignore. People do not burn out because the meditation app was insufficient. They burn out because the way work is led is unsustainable.

This is not a soft observation. It is a structural one. If the leader is the largest single variable in team engagement, and twenty years of Gallup data suggests they are, then wellbeing has to be treated as a leadership output, not an HR programme.

 

Six disciplines of a healthy leader

The shorthand I have come to use is L.E.A.D.E.R (an obvious one). Six disciplines that the leaders I have watched succeed over a long career practise without making a virtue of it, and the leaders I have watched burn out, or burn others out, tend to neglect at least three of them.

 

L – Limits

The hardest to practise and the most easily faked. A healthy leader holds boundaries on their own working hours, their decision-making capacity, and the volume of work they will absorb before pushing back. The leader who answers every message at eleven at night signals to the organisation that the line is somewhere past eleven at night, regardless of what the wellbeing policy says. Limits are taught by example or they are not taught at all.

E – Empathy

Not the soft listening that fills articles about emotional intelligence. Disciplined empathy. The ability to read what is happening in a room, to notice the team member who has stopped speaking up, to interpret a steering committee mood before reacting to the slide. Empathy without standards is unprofessional. Standards without empathy are unsustainable. Both at once is the discipline.

A – Accountability

Healthy leaders take more responsibility than the role formally requires. They own the call, they own the consequence, and they correct themselves in public when the call was wrong. The cultural rot in most organisations is not weak performance. It is leaders who quietly redirect accountability downwards when results disappoint. People watch for this. Once they see it, the relationship is over.

D – Discipline

The daily habits that produce sustained executive performance over a thirty-year career rather than a brilliant five-year sprint. Three priorities written down. A structured one-to-one rhythm. Time blocked for thinking. A weekly review. Discipline is not glamorous and it is not strategic. It is the unglamorous, unstrategic foundation that everything else stands on.

E – Energy

A leader is, among other things, a capacity manager. Their own capacity, and the capacity of the people around them. The healthy leader treats sleep, recovery, and physical condition as professional obligations, not personal preferences. They notice when the team is running on reserves and adjust the operating tempo before something breaks. The unhealthy leader treats exhaustion as a badge of seriousness and accidentally institutionalises burnout as a sign of commitment.

R – Reflect

The discipline most often dropped first under pressure, and the one that most reliably separates leaders who compound over time from those who plateau. Healthy leaders make space to ask what worked, what did not, and what the next iteration looks like. They do this weekly, not annually. They write it down. It costs them an hour a week and it compounds over a career.

 

Wellbeing was always the leader’s job

If you read the six pillars and notice that none of them are particularly new, that is the point. There is no insight here that has not been written about for thirty years. The insight is in the arrangement. Six things, practised together, by the leader. Not delegated to HR. Not outsourced to a vendor. Not performed on stage at the annual offsite.

A wellbeing programme without these six behaviours is a sticking plaster on a system designed to bleed. A leader who practises these six things, in a company with no formal wellbeing programme at all, will produce a healthier organisation than the alternative.

The temptation when reading frameworks like this is to nod, save the post, and carry on operating exactly as before. The honest test is the diary. If your week, the way you actually spend the hours, does not reflect at least four of these six disciplines, the wellbeing of the people who work for you is already on a slow countdown, regardless of what your engagement survey says.

There is no other wellbeing programme. Only how the leader spends the week.

Pre-Mortem: Anthropic’s Wall Street Agentic AI Suite

Posted on by

 

Thirteen of the world’s largest financial institutions just deployed ten autonomous AI agents into the most regulated workflows in finance. None of them has publicly named who is accountable when the agents are wrong. Not the banks. Not the vendor. Not the regulators. The launch on 5 May reads like a milestone. Read closer and it reads like a stress test of every governance assumption the financial services industry operates on.

A post-mortem tells you why something failed once it already has. A pre-mortem asks the same questions before failure is possible. Same five questions, every time, applied to a current programme, announcement, or initiative. This is the first in the series, and the subject is not chosen by accident. The Anthropic Wall Street launch is the clearest example I have seen this year of capability racing ahead of the architecture meant to hold it to account. If you are a CIO, a CRO, or a transformation lead in a regulated industry, the lessons here apply to you whether you are deploying Claude or not.

 

The Bet

Anthropic and the deploying banks are betting that ten autonomous agents can land in the most regulated workflows in finance, underwriting, KYC, credit memos, statement audits, faster than the regulatory architecture can constrain them. The technical bet rides on Claude Opus 4.7’s 64.37% on the Vals AI Finance Agent benchmark and AIG’s quoted 88% accuracy on insurance claims out of the box. The strategic bet is that being first at this footprint, including JPMorgan Chase, Goldman Sachs, Citi, AIG, BNY, Carlyle, Mizuho, and Visa, outweighs whatever comes back from regulators in the next twelve months. Reasoned bets, made by an extraordinarily capable vendor and the most sophisticated buyers in the world. But they are bets, not certainties, and the launch reads as certainty. The CIO of any one of those banks is taking on operational, regulatory, and reputational risk for which the vendor has accepted no published share. That is the bet they should be examining most carefully.

 

The Assumption

One belief is doing all the work: that bank operating models can absorb ten simultaneously deployed agents without the human-in-the-loop quietly thinning where the agents prove reliable. Anthropic’s own commitment depends on it, from the primary announcement: “Users stay firmly in the loop, reviewing, iterating on, and approving Claude’s work before it goes to a client, gets filed, or is acted on.” The history of automation in regulated environments tells a different story. Algorithmic trading kill switches were not triggered because the system was performing. Automated underwriting reviews became rubber stamps once approval rates looked normal. Every automation failure in regulated finance follows the same arc: human oversight erodes invisibly as the system proves itself, and the erosion is only visible after the failure. JPMorgan CIO Lori Beer said it directly at the launch: “The technology can do so much. It’s the actual organization’s ability to digest and absorb it.” That ability is the load-bearing assumption. If it holds, the launch is a milestone. If it does not, the launch is a slow-moving incident.

 

The Sequence

Capability shipped. Ten named agents, Microsoft 365 generally available, Moody’s embedded, more than a dozen banks in production. What was committed before the operational governance for vendor-supplied agentic decisioning was published: all of it. Three weeks earlier, the Fed and the OCC revised Model Risk Management guidance and explicitly excluded agentic AI as “novel and rapidly evolving.” A Request for Information is planned, with no committed timeline. The EU AI Act’s high-risk financial-sector requirements take effect 2 August, twelve weeks after launch. The FCA and PRA decided against creating a dedicated AI Senior Management Function and instead mapped accountability onto existing SMFs that were never designed with autonomous agents in mind. Three jurisdictions. Three different gaps. One vendor launch landing in all of them at once. This is not a regulator being slow. This is a regulator explicitly stating that the rules do not yet apply, while the systems the rules are meant to govern are already in production.

 

The Pager

The banks have named regulatory accountability at the firm level. SMF24 (Chief Operations), SMF4 (Chief Risk Officer), SMF16 (Compliance Oversight) at FCA and PRA-regulated firms hold statutory responsibility for technology, risk, and compliance. Model risk owners at US firm level cover the same ground. Real, senior, public. That deserves credit. However, none of them have been publicly named for the deployment of these specific agents. Inheriting accountability through a job description is not the same as being named as the accountable owner of a programme. The first is the regulatory default. The second is what serious AI governance actually requires. Anthropic has no published vendor accountability commitment for autonomous regulated decisioning. The asymmetry is the entire story. When a Claude-built agent denies a loan that should have been approved, or approves a KYC file that should have been escalated, the pager rings at the bank, with consequences for the bank, while the vendor’s exposure is contractual and capped. The clearest demonstration came six days before the launch itself. On 29 April, Goldman Sachs removed Claude access for its Hong Kong bankers over contractual, regulatory, and geopolitical factors. The bank pulled the product. The vendor did not pull itself out. Whoever absorbs the cost when regulatory fit fails, absorbs it alone. Until vendor accountability is publicly framed, every bank deploying these agents is underwriting risk the vendor will not.

 

The Proof

Two outcome measures have been published. 64.37% on Vals AI. 88% on AIG insurance claims out of the box. Both are useful. Neither measures regulated-decision accuracy at scale. There is no committed measure for customer-detriment rate, near-miss frequency, incident reporting cadence to regulators, or the rate at which human reviewers actually amend agent outputs versus rubber-stamp them. The banks deploying these agents do not yet have public outcome commitments either, and that absence is its own answer. Former CFO Alyona Mysko captured what is at stake: “In finance, 99% correct is still wrong.” In eighteen months, the question “did this work?” will be answered by whoever owns the platform to define what work means. Right now, that platform is the vendor’s marketing. The banks need to claim that platform back, in their own outcome language, before the metric is set by a third party with no skin in their game.

 

Verdict

The launch is genuinely significant. More than a dozen named banks in production, industry-leading benchmark performance, audit logs in the Claude Console, the deepest Microsoft and Moody’s integrations any AI vendor has shipped. None of that is in dispute.

What is in dispute is whether the deploying banks have done the work to fill the accountability gap that the vendor has not closed and the regulators have not yet defined. The lesson generalises beyond Anthropic and beyond banking. Any CIO buying agentic AI in a regulated industry, healthcare, insurance, energy, the public sector, is operating in the same gap, and most have not yet noticed.

The action is concrete. Name the human in your organisation who carries the pager when the agent is wrong. Demand a vendor accountability schedule before you sign, not after. Define your own regulated-decision outcome measure and publish it, so the standard your performance is judged against is one you helped set.

If Anthropic publishes a vendor accountability commitment in the next six months, and a major bank commits to a public regulated-decision outcome measure tied to a named owner, this becomes a case study other industries will study for years. Without both, it becomes the most expensive procurement lesson the industry buys this decade.

The UAE Leads the World in AI Adoption. That Is the Easy Part

Posted on by

 

The UAE’s 70% AI adoption figure is everywhere. Conference keynotes open with it. Board papers cite it. Technology leaders in the region are being measured against it.

It is an impressive number. The UAE leads the world, ahead of a global average of just 17.8%. Government entities are reporting 97% AI tool adoption. Investment in AI infrastructure exceeded AED 543 billion across 2024 and 2025. The commitment is real, it is visible, and it is serious.

But a figure that measures how many people are using AI tools does not tell you whether those tools are being used well, safely, or in ways that will actually compound into competitive advantage. Right now, across the region, adoption has outpaced governance, capability, and leadership readiness by a significant margin.

That is the conversation worth having.

 

The Gap Between Using and Doing Well

A Fast Company Middle East report found that skills gaps, governance issues, and resource shortages are actively hindering AI projects across the UAE and Saudi Arabia. McKinsey found 88% of organisations globally now use AI in at least one business function. In a separate McKinsey study, only 1% of leaders called their own AI deployment mature.

Read that again. 88% usage. 1% maturity.

Most of the adoption conversation is measuring the first number. Almost nobody, anywhere, is meaningfully achieving the second.

This is not a reason for pessimism. It is a reason for precision. Because the organisations that close that gap are the ones that will extract genuine long-term value from the investments being made. The ones that do not will have impressive statistics and quietly disappointing outcomes.

 

What the 70% Figure Actually Measures

Adoption, in most surveys, means someone in the organisation is using an AI tool. It does not mean:

  • Those tools are connected to meaningful business outcomes
  • There is a governance framework determining how AI agents operate, with what access, and under what oversight
  • Leaders understand the capability well enough to ask the right questions of it
  • The organisation has redesigned workflows around AI rather than simply layered it on top of existing ones
  • There is a plan for what happens when something goes wrong

“Adoption measures presence, not performance. A Copilot licence in every seat is not a transformation. It is a starting point.”

 

 

What Sits Underneath the Headline Number

The organisations that will genuinely lead in this environment are not the ones chasing the adoption number. They are the ones building what sits underneath it.

Three things separate the organisations that will compound this investment from the ones that will stall.

  • Governance before scale. As AI agents take on more autonomous tasks, the permission architecture, oversight mechanisms, and human confirmation requirements need to be established before deployment at scale, not retrofitted after something goes wrong. Look at the incidents of the last eighteen months. Production databases deleted. Cloud environments wiped in seconds. All of it the consequence of deploying capability ahead of governance.
  • Leadership readiness, not just technology literacy. Most AI adoption programs focus on upskilling employees to use tools. Far fewer focus on equipping leaders to make good decisions about AI: what to deploy, what oversight to maintain, what risks to accept, and what questions to ask the vendors selling them the infrastructure. “Technology literacy and leadership readiness are not the same thing.” Confusing the two is one of the most common and costly mistakes being made right now.
  • Workflow redesign, not workflow overlay. The organisations getting lasting value are not the ones that added AI to existing processes. They are the ones that redesigned the process around what AI can actually do. That requires change management discipline, not just technology deployment.

 

The Region Has the Ambition. Now It Needs the Architecture.

The UAE’s strategic commitment to AI is not in question. A 543AED billion investment, a world-first framework to deploy agentic AI across government, a national curriculum introducing AI literacy from school level. These are not the moves of an economy dabbling. This is a serious long-term play.

That is exactly why the governance and capability conversation matters so much right now. The investment is in place. The infrastructure is being built. The adoption numbers are world-leading.

The question is not whether the UAE is committed to AI leadership. It clearly is. The question is whether the organisations operating within that environment are building the internal foundations to convert the headline numbers into durable, compounding advantage.

A 70% adoption rate is the beginning of the story, not the destination.

 

The Organisations That Will Lead Are Already Asking Different Questions

They are not asking how to get their adoption rate up.

They are asking what good looks like once they get there. Who is accountable for how their AI agents behave. What their governance architecture looks like for the autonomous systems they are deploying. What they are actually measuring to know this is working.

Those organisations will not be the loudest voices at the next conference. They will be the ones with something real to show for it in three years.

The UAE’s 70% AI adoption figure is everywhere right now. It is genuinely world-leading, and it is not the number that should be keeping leaders awake at night.

Globally, 88% of organisations use AI. 1% have reached actual maturity. That is the gap worth talking about, and the organisations closing it are not the ones chasing higher adoption rates.

The question I keep coming back to: if your organisation is sitting in that 70%, who is actually accountable for how your AI agents behave once they are deployed?