Hi friends,

Goldman Sachs' 2026 summer interns were told to experiment with AI from day one.

It's an interesting signal about how AI is entering financial services.

Not because every graduate is suddenly AI-native. They aren't.

But some of the newest employees entering the industry are arriving with a willingness to experiment with tools that didn't exist when many of their managers started their careers.

And they're entering organisations where everyone is being asked to experiment.

Managers included.

At the same time, the layer between the executive team and the front line is under pressure.

Gartner predicts that by the end of this year, 20% of organisations will use AI to flatten their structures, in what has been dubbed the Great Flattening, with more than half of current middle-management positions set to be eliminated.

Bloomberg Intelligence estimates that global banks could cut as many as 200,000 jobs over the next three to five years, with back-office, middle-office and operations among the areas most exposed.

So there is a strange tension emerging.

The organisation is asking middle managers to do more with AI at precisely the moment it is asking whether it needs as many middle managers at all.

AI-generated for The Gaziza Lens

And "use AI" is not a single new responsibility.

The middle manager is now expected to experiment with new tools, work out which ones are actually useful rather than just impressive, redesign the workflow around what works, help their team adapt without losing sight of where human judgement still matters, and turn all of that into something the business can measure. That's five different jobs stacked on top of the one they already had.

All while keeping the existing operation running.

That is a much bigger job than simply knowing how to use ChatGPT.

And it creates an interesting organisational dynamic.

An early-career employee might discover a new AI workflow.

A senior leader might set an ambitious productivity target.

But the middle manager is often the person who has to work out: does this actually work? can we use it safely? and who is accountable when it gets it wrong?

This is where I think the middle-management story becomes more interesting.

The issue isn't that AI-native juniors will replace experienced managers.

It's that AI is changing what the organisation expects from the manager in between.

The old management job was largely about allocating work, developing people and monitoring output.

The emerging job is broader: experiment with AI, manage people using AI, redesign workflows, make decisions, manage the risks, own the outcome.

The irony is that the manager may have less time than almost anyone else to figure all of this out.

The junior may have time to experiment.

The executive has the authority to set the direction.

The middle manager has the accountability and the existing workload.

So perhaps the question isn't whether middle managers disappear.

It's:

What happens when we keep the accountability but fundamentally change the work?

Because if AI removes some of the coordination work that created middle-management roles in the first place, organisations will need to be much more deliberate about what they want the manager to do instead.

The future middle manager may not simply be someone who manages people.

They may be the person who can turn AI experimentation into a better way of working, while knowing where not to automate, what to change, and what they remain personally accountable for.

That's a different kind of management job.

And financial services is going to need to figure it out.

If you do one thing before the next issue

For this one:

Look at your own role and ask: “If I had to deliver the same outcomes with half the headcount, what would I redesign first?”

Then ask:

“What would I stop doing altogether?”

Food for thought.

Sources: Goldman Sachs, "Welcome to Our 2026 Summer Interns" · Gartner, "Top Predictions for IT Organizations and Users in 2025 and Beyond" · Bloomberg, "Wall Street Expected to Shed 200,000 Jobs as AI Erodes Roles"

Thanks for reading. If this one landed for you, sharing it with someone who'd find it useful means a lot.

Recommended for you

View all
caret-right