Emily Sauter, Head of AI Organisation

Make no little plans; they have no magic to stir men’s blood.” Daniel Burnham

Where the first paper left off

The first paper in this series argued that the constraint on AI value is organisational rather than technological, and that the organisations drawing down their reserves of human judgement while dismantling the structures that replenish it are accruing a ‘judgement debt’ that falls due years after the executives who incurred it have moved on.

It closed by recommending that you first establish where your organisation sits on AI maturity, where you intend to get to, and how fast. This paper sets out how to progress, what it means for HR, and how it shapes your organisation.

Daniel Hudson Burnham (1846-1912) was one of the most influential architects and urban planners in American history, and arguably the person most responsible for the modern vision of Chicago.

Chicago is a city that stole my heart, and I spent many of my happiest days near the Chicago River. The bridges that cross its waters, the boats that traverse its curves, and one time a year, it turns green for the merriment of St. Patrick Day.

But it wasn’t always a place of beauty and charm.

In the late nineteenth century, Chicago had an increasingly dangerous problem. The city was growing rapidly, but the systems beneath it had not kept pace. The Chicago River carried human and industrial waste into Lake Michigan, the same lake from which the city drew its drinking water.

The obvious solutions were incremental: move the water intakes farther offshore, improve the pumps, manage the symptoms. Chicago eventually chose something more radical. It changed the direction of the river.

To do so, the city needed more than an engineering idea. It needed a new institution with the authority to act, years of excavation, new infrastructure, sustained investment and a willingness to challenge what had previously seemed fixed. In January 1900, water began flowing away from Lake Michigan and towards the Des Plaines River and the Mississippi River watershed. [mwrd.org], [wttw.com]

The breakthrough was not simply a better piece of machinery. It was the redesign of an entire system.

That distinction matters now.

Many organisations are approaching artificial intelligence as Chicago might have approached its polluted river by buying stronger pumps. They are adding copilots to existing roles, automating fragments of existing processes and measuring how often people use the tools. Each intervention may be useful. None, on its own, changes the direction of the organisation.

Five levels, and what defines them (as seen in Part 1)

3H (Head-Heart-Hands) AI Maturity Model (© Bendelta, 2026)

Bendelta’s AI Maturity Model describes five levels. At Level 1, AI assists work — it acts as a smart assistant or copilot. At Level 2 it enhances work, overseeing and improving it. At Level 3 it automates work, operating on its own and doing more with less. At Level 4 it reinvents work, enabling a whole new way of operating. At Level 5 it invents work, driving new products and services that did not previously exist.

The distinction that matters is this: a level is defined by how far the work has been reimagined, not by which tools have been purchased. An organisation with every licence in the catalogue and unchanged workflows is at Level 1. There’s no better way to waste money on tokens, then to invest in technology without intentionally redesigning work.

The distribution is sobering or comforting, depending on your perspective. According to the data from our Chief People Officer Roundtables, 83% of our clients place their organisation at L1 to L2. Only 17% are at L3 or above. And more than half want to genuinely reinvent work and create new forms of value (36% aspire to level 4 and 19% for level 5).

Where’s the natural landing point for where organisations are, today? Level 3. 44% of Chief People Officers nominated level 3 as the aim point, which suggests automation is the next practical step, even while a significant group is already thinking beyond automation to reinvention.

Three rules govern the current

1. Not every organisation should be aiming for Level 5

How far you go depends on the readiness of your workforce, the state of disruption in your market, your tolerance for risk and your appetite for investment. A Level 3 organisation that is genuinely, deeply at Level 3 will outperform a Level 4 organisation held together by hope.

2. Capability is built level by level

Each level puts in place the building blocks that make the next one possible. Jumping ahead is usually impossible to do well and always wastes money. Organisations attempt it constantly — jump to automating processes at Level 3 without the capabilities of Level 1 and 2. Think failed pilots, quiet workarounds and a workforce that has learned not to trust the next announcement.

3. Parts of your organisation will sit at different levels, and that is normal

Early adopters mature faster; other functions lag. This is not a problem to be eliminated. It is a condition to be managed, and it requires deliberate practice rather than a single enterprise-wide program. This also means business leaders need to consider the specific needs of their part of the organisation and meet them where they are.

Who directs the river (workflow)

In technology transformation, we’ve always questioned who owns it: business or technology leadership. The decision will determine the results. But if we agree the real value happens in work transformation, not artificial intelligence alone – there’s a player we’ve often undervalued who is the best fit to lead.

The Chief People Officer understands how humans work, what they need to be successful, and the architecture of data that determines their success. As organisations progress in the maturity curve, the information that binds the organisation begins to change.

At Level 1, tasks are executed differently. At Level 2, the tasks themselves change and roles and responsibilities evolve. At Level 3, work is outsourced to agents, and team design adjusts to treat agents like team members. At Level 4, organisation design changes rapidly, and only at Level 5 do you finally understand how the operating model itself becomes something new.

The Chief People Officer understands this architecture. They understand how humans are motivated and driven to work. Therefore, the role they play is imperative.

Often in AI transformation, technology leads the conversation but once you arrive at Level 3, agents and humans are working together. You’re becoming “agentic.” And no one knows how to identify, onboard, train, and manage performance of a resource better than People and Culture. Hand them the technology capability and architecture, and they can now identify, onboard, train, and manage performance of your agents, too. Together, you can start to enter Level 4 and Level 5.

The mistake that most organisations have made is giving people “free will” to create agents as they see fit. You’d never hire heaps of humans without careful consideration – and the same is true for agents. Agents need onboarding, permissions, performance standards, escalation paths and eventually decommissioning. They make errors that look like conduct issues. Ask who in your organisation is accountable for that population today, and the honest answer is usually either ‘nobody’ or ‘whoever bought the licence’.

The functions that will own it are the ones that build the capability to own it before the question is asked. Most HR functions are not currently built for this, and the gap is not one of intent. It’s clear direction on how far, and how fast, the river should flow.

Who is standing at the riverbank

There is another important figure in this story: the engineer standing at the riverbank. They can see where value is created, where effort is wasted, which processes frustrate people, where judgement must remain human, where teams are ready to experiment, and where leaders may be creating resistance without realising it.

While the Chief People Officer decides if the river should change direction, the engineer or HR Business Partner decides where the river flows.

They see the informal processes, hidden dependencies and recurring frustrations that operating models rarely capture. As organisations progress from Assist to Reinvent, the challenge becomes less about deploying technology and more about redesigning work itself. HR Business Partners are uniquely positioned to identify where that redesign should occur. In many organisations, they are the bridge between strategic aspiration and operational reality. They understand how decisions move through the organisation, where knowledge becomes trapped, where effort is duplicated and where workarounds have become normal practice. That insight becomes essential when determining where AI should assist, enhance, automate or fundamentally reinvent work.

HR Business Partners already specialise in workforce planning, talent, engagement, performance, and change. Arm them with the engineering skills to reimagine work, and the authority to decide or inform how work should be done, and now we are creating real momentum. You’re solving the biggest hurdle of all: how to capture organisation design, workforce strategy, and business performance to win in the era of artificial intelligence.

Deciding how far, and how fast

Considering that AI Maturity is a human and organisational challenge as much as a technology one, we asked our Chief People Officers at the CPO Roundtable what’s holding them back.

28% mentioned capability, skills, confidence, and adoption. 26% mentioned identifying value, clear use cases, strategic alignment, and intentionality. Around 19% mentioned trust, risk appetite, patient safety, ethics, governance, or accountability. 13% mentioned legacy technology, poor foundations, and data accuracy; while 17% mentioned multi-generational workforces, relationships, and concerns about removing the human element.

How can organisations change the current with such substantial hurdles in the way?

Your target is not set by what a competitor announced at a conference or by what you’re reading on LinkedIn. It is being thoughtful and intentional about how you answer these seven questions, and they are worth asking explicitly and out loud, with the executive team in the room:

• Risk – How much tolerance do we have to take risks?
• Speed – How quickly are we prepared to adapt and change?
• Customer – How much can we reimagine the service we provide?
• Operational – How much opportunity do we have to reimagine the way we operate?
• Employees – How much do we involve our people?
• Technology – How adaptable is our data and technology stack?
• Spend – What level of investment are we genuinely willing to make?

The seventh answer usually contradicts the first six. That contradiction is the most valuable thing the exercise produces, and it is better discovered in a room than in a post-implementation review. The good news is that intentionality usually leads to much better return on investment.

The way the system flows

The same way the Chicago River is essential to the function of my favourite city, Chicago – work is essential to the value you’re hoping to achieve through artificial intelligence.

Throwing more tools, more pilots, and more training at your people won’t lead them to improve how they work. The key is to work directly with your workforce to know the work, analyse the work, and reinvent the work. Our next white paper will discuss exactly this: how do you know where humans are still the best intelligence to get the job done, and where they can accelerate outcomes by using artificial intelligence.

The reversal of the Chicago River wasn’t easy, and it was met with a lot of resistance. Communities downstream challenged the decision. Our next paper will share how to address the Heads, Hearts, and Hands of your workforce to alleviate resistance in the system.

The challenge is that a river, or an organisation, cannot be redirected by ambition or technology alone.

You do not change the current by moving faster in the same direction.

The Bendelta AI Maturity Model © Bendelta 2026.
Bendelta client experience across industries; Burnham,
D.H. (1910). The Development of Cities of the Future. Metropolitan Water Reclamation District of Greater Chicago. (2025). Chicago River’s reversal in 1900 was an engineering triumph that transformed our city. WTTW Chicago. How Chicago Reversed Its River: An Animated History.