A year after IBM’s Deep Blue defeated Garry Kasparov in 1997, Kasparov did something unexpected. Rather than retreat, he invented a new form of chess he called ‘advanced chess’, pairing human players with computers to see what they could produce together. The result was remarkable. Even moderately skilled players, armed with a standard machine, were capable of defeating both grandmasters playing alone and computers operating without human input. The combination was categorically superior to either element in isolation.

By: David Henderson , Group CHRO, Al-Futtaim 

That experiment carries an important lesson for organisations navigating AI today. The instinct understandable, but mistaken, is to frame AI as a technology story. It is not. AI reshapes jobs, redistributes decision rights, resets operating models, and forces us to reconsider deeply embedded ways of working. It intersects directly with creativity, cognition, confidence, identity and employability. It produces as many human questions as it does technical ones.

This is why the organizations that are genuinely converting AI from experiment into competitive advantage are those that have understood it, first and foremost, as a large-scale human transformation, one that demands the business, the CHRO and the CIO working as genuine partners, each bringing what the other cannot.

The organisations winning with AI are not those with the most sophisticated technology. They are those that have most deliberately redesigned how humans and machines work together.

The Case for the CHRO

The most effective AI transformations are driven by a tight three-way partnership:

the business setting the agenda and owning outcomes,

the CIO providing the technology platforms,

data infrastructure and governance,

and the CHRO leading the human transformation that determines whether AI delivers value at scale or stalls in pilots.

Each is essential. None is sufficient alone.

What has changed is the recognition that the human dimension, the design of work and decision rights, the building of workforce capability, the management of trust and ethics, the orchestration of adoption across large and diverse employee populations, is not downstream of the technology. It is a primary enabler of it. That is the CHRO’s territory, and it demands the same strategic weight as the technology agenda itself.

In this paper, I propose a model for how CHROs can lead AI enablement through four interconnected roles: Design Architect, Capability Steward, Adoption Catalyst, and Transition Guardian. Each role addresses a distinct dimension of the human transformation that AI demands. Together, they represent a holistic operating mandate for CHROs who are serious about delivering sustained enterprise value from AI, not just deploying tools.

01) Design Architect: Redesigning work, roles and decision rights for the AI era

AI transformation fails far more often because of organisational design choices than because of technology limitations. When companies deploy AI tools without redesigning how work is done, decision rights blur, accountability erodes, adoption stalls, and productivity gains remain trapped in pilots. The technology is rarely the binding constraint. The organisation almost always is.

The CHRO’s role as Design Architect is to get ahead of that problem. This means providing overarching direction on how work should be redesigned so that human judgment and AI-generated insight are deliberately combined, not accidentally layered on top of each other. It means clarifying which decisions remain human-led, which are AI-supported, and where accountability ultimately sits. And it means building an operating model architecture that is dynamic enough to evolve as AI capabilities continue to develop rapidly.

In my own experience, incrementalism in this domain is almost always destined to fail. The organisations that are getting this right are making bold, decisive design choices, and in some cases, breaking up parts of the organisation that have long been treated as untouchable.

In Practice — Procter & Gamble P&G redesigned decision models across forecasting, procurement and product innovation so that AI produces insights and options while humans retain final say on portfolio bets, supplier strategy and innovation priorities. Critically, AI was embedded directly into logistics decision forums — rather than remaining siloed in group-level analytics teams, removing information-sharing barriers and enabling real-time decision-making at scale.

In Practice — Microsoft Microsoft intentionally redesigned all knowledge-work roles so that AI copilots handle drafting, synthesis and retrieval, while employees retain judgment, prioritisation and accountability. The result was not simply cost reduction,it was the redeployment of released cognitive capacity into revenue-generating innovation and customer experience improvement.

Being intentional on organisational design means staying one step ahead of technological adoption, not one step behind it. The CHRO must proactively reimagine how AI reshapes the value chain and translate that vision into operating model decisions — rather than reactively course-correcting after tools have already been deployed.

02) Capability Steward: Building enterprise-wide, continuous learning systems that keep pace with AI

In the AI era, capability, not technology, is the primary constraint on value creation. The organisations that are scaling AI effectively are not those with the most sophisticated tools. They are those whose people know how to use them confidently, critically, and productively in the context of real work.

The CHRO’s role as Capability Steward is to build the learning infrastructure that makes this possible at scale. This means moving decisively away from episodic, one-size-fits-all training models, which are structurally unsuited to the pace of AI change, towards continuous, contextual learning systems that are embedded in daily workflows.

It means developing AI fluency across the workforce, not just in specialist teams. And it means maintaining ongoing insight into which capabilities are emerging, shifting or declining as the skills economy evolves.

In Practice — Amazon Amazon treats AI capability as core workforce infrastructure rather than a specialist skill. It has built role-specific learning pathways combining foundational AI fluency with immediate, in-role application, particularly in operations, logistics and corporate functions. The result has been faster adoption of AI tools across large frontline and corporate populations, with measurable productivity gains driven by applied capability rather than isolated expertise.

From My Experience — Zurich Insurance During my time at Zurich, we built an enterprise-wide AI and digital capability ecosystem that combined broad AI literacy with deep domain-specific learning for underwriters, claims handlers and risk professionals. Learning was continuous and embedded in daily workflows. Critically, we also focused on transferable skill identification, enabling us, for example, to rapidly retrain and redeploy claims handlers as customer service agents based on strong overlaps in their underlying skill profiles. That flexibility became a genuine competitive asset.

The CHRO must protect long-term capability health and resilience, not simply optimise for short-term productivity. Organisations that treat AI learning as a one-time training event will struggle to sustain adoption. Those that build continuous learning as an organisational capability will compound their advantage over time.

03) Adoption Catalyst: Empowering employees as co-creators of AI value, not passive recipients of it

Many CHROs of my generation were trained in a change management orthodoxy that starts at the top of the house, guiding coalition, executive sponsorship, structured project timelines. That model is not wrong, but it is increasingly insufficient for AI.

Top-down governance and strategy remain essential. But scalable AI value does not come from mandates. It comes from the bottom up, from employees who understand the work and are empowered to apply AI where insight is deepest and value most immediate.

The CHRO’s role as Adoption Catalyst is to create the conditions for this to happen: building cultures of experimentation and knowledge-sharing, aligning incentives and recognition to reward participation, and enabling employees to co-create AI use cases rather than simply receive them.

This is a fundamental shift from change management to what I would call change orchestration, leaders creating the environment in which adoption flourishes, rather than driving it through compliance.

In Practice — Al-Futtaim Blue Loyalty Platform The clearest proof point I can offer comes from our own experience at Al-Futtaim. The group’s Blue Loyalty Platform uses AI to combine behavioural, transactional and partner data to deliver personalised offers and purchase recommendations across our retail and service channels. What made this work was not central design — it was that the use cases were developed by multi-disciplinary frontline retail employees, working in agile action-learning teams, applying their direct customer insight to build the recommendations. AI was embedded into frontline and digital workflows by the people who understood those workflows best. The result has been measurable revenue uplift driven by use cases rooted in real customer interactions — not boardroom hypotheses.

In Practice — Google Google runs AI adoption through a culture of experimentation supported by internal communities, shared tooling and lightweight governance. Employees apply AI to improve workflows, products and services; successful use cases are productised and scaled through internal platforms. This produces rapid diffusion of best practices, strong employee ownership, and continuous improvement generated by those doing the work.

Employees need to define the tools they need , not simply learn the tools they are given. That distinction is everything when it comes to whether AI adoption takes root or stalls.

Bottom-up adoption is not a cultural nicety. It is the mechanism through which AI becomes embedded, differentiated and commercially meaningful at scale. Organisations that get this right do not deploy AI. They make AI part of how the organisation thinks.

04) Transition Guardian: Ensuring AI adoption is ethical, transparent, and in the long-term interest of employees

AI introduces legitimate concerns that the CHRO cannot afford to minimise: fairness, transparency, surveillance, bias, job security, long-term employability. If these concerns are not addressed proactively and honestly, trust erodes, and without trust, adoption stalls regardless of how good the technology is.

The CHRO’s role as Transition Guardian is to ensure that AI adoption is consistent with organisational values and strengthens, rather than undermines, the employee value proposition.

This means embedding ethical guardrails and human oversight into AI adoption from the outset, not retrofitting them under regulatory pressure. It means communicating honestly with employees about what AI will change, what it will not change, and what pathways exist for reskilling and redeployment.

And it means treating strategic workforce planning not as an HR administrative function, but as a core enabler of organisational resilience.

Today’s employees need to focus less on specific target jobs and more on building transferable skill profiles that will serve them across a career that is certain to be turbulent. They need to feel that their organisation has their back. The CHRO must make that commitment credible, not through reassurance, but through concrete pathways.

In Practice — Salesforce Salesforce has embedded ethical and responsible AI as a prerequisite for scale rather than a control imposed after deployment. The company requires mandatory Responsible AI training, applies humanin-the-loop oversight for AI-enabled decisions, and maintains clear disclosure standards when AI influences employee or customer outcomes. The trust this generates has driven faster adoption, stronger employee engagement, and meaningfully reduced legal, regulatory and reputational risk.

In Practice — Unilever Unilever explicitly links AI adoption to employability and internal mobility. As AI reshapes roles, the company invests heavily in reskilling and redeployment pathways, reframing AI as augmentation rather than displacement. Workforce planning, learning and ethics are intentionally connected rather than siloed , and employees can see a credible future for themselves within the transformation.

Trust is not a soft outcome of AI transformation. It is the hard prerequisite for scaling it. The CHRO who treats it as such will find that ethical, transparent AI adoption does not slow the transformation down — it is the thing that makes it durable.

The CHRO Skill set for AI Enablement

Having defined the four roles the CHRO must play, it is worth being specific about the skills and attributes required to execute each one. In an environment where AI success is increasingly determined by organisational design, capability building, adoption dynamics and trust, not technology, these capabilities define whether the CHRO is shaping the transformation or reacting to it.

Design Architect	Capability Steward	Adoption Catalyst	Transition Guardian
Operating Model Design	Learning at Scale	Change Orchestration	Ethical Judgement
Work & Role Deconstruction	AI Fluency Translation	Employee Empowerment Mindset	Trust Stewardship
Decision Rights Clarity	Skills Architecture & Workforce Sensing	Incentive & Recognition Design	Strategic Workforce Planning
Systems Thinking	Action Learning Systems	Business Experimentation Literacy	Risk Anticipation
Enterprise Co-Creation	Future Capability Stewardship	Cultural Signal Awareness	Clear, Honest Communication

A few points of emphasis.

As Design Architect, the most underrated skill is enterprise co-creation — the confidence and credibility to act as a genuine co-owner of AI strategy with the CIO and business leaders, not merely as a supporting function.

As Capability Steward, future capability stewardship is distinct from short-term productivity optimization; CHROs must protect long-term organisational resilience, not just near-term performance.

As Adoption Catalyst, cultural signal awareness is often more powerful than formal programmes, leadership language and behaviour either accelerate or silently undermine adoption at scale. And as Transition Guardian, clear and honest communication, including on uncertainty and difficult tradeoffs, is the foundation on which all of the other skills rest.

Without it, none of the others land.

Conclusion: The Human Transformation Imperative

Organisations that are genuinely winning with AI are not those with the most sophisticated technology stacks. They are those that have most deliberately and thoughtfully redesigned how humans and machines work together, rethinking operating models, building capability at scale, empowering employees as co-creators, and managing the transition with ethics and transparency.

The CHRO who grasps this, who acts as Design Architect, Capability Steward, Adoption Catalyst and Transition Guardian simultaneously, becomes one of the most important executives in the organisation. Not because HR has staked a claim to a technology agenda, but because the most important levers for AI value creation are organisational and human, and those are precisely the levers that CHROs are equipped to pull.

Kasparov’s advanced chess experiment showed us, a quarter of a century ago, that the most powerful outcomes emerge not from humans or machines working alone, but from their deliberate, skillful combination. The CHRO’s mandate is to make that combination work, at enterprise scale, at pace, and without losing the trust of the people it depends on.

That is not a supporting role. It is a defining one.

_______________________________________________________

David Henderson is Group CHRO of Al-Futtaim Group, one of the Middle East's largest diversified conglomerates. He has previously served as CHRO of Zurich Insurance Group, MetLife and PepsiCo.

Christiana Maxion, Founder and CEO of MAXION

Consumer apps have never been easier to use. With AI improving navigation, personalization, and responsiveness, platforms now offer a far more seamless experience, helping users move through tasks, content, and decisions with little visible effort. But convenience alone is not the same as value. Recent research found that the average adult now spends 88 days a year on their phone, highlighting both the scale of digital dependence and the urgency of building products that deliver something more meaningful than another scroll session.

Concurrently, expectations have changed. McKinsey has reported that 71% of consumers expect personalized interactions, while KPMG’s UAE research shows that integrity has now overtaken personalization as the strongest driver of customer experience. People still want services that understand them, but they also want trust and clarity that technology is working in their interest.

This is the backdrop for the rise of behavioural AI in consumer apps. The next phase of app design will be judged by its ability to predict what a user may click next, and more by how well it turns intent into action with less friction.

The problem with designing for activity, not action

For years, most consumer platforms have optimized for clicks, scroll depth, watch time, and repeat visits. Those metrics are useful, but incomplete. They show that a user remained active, not whether the user made progress.

A person may spend 20 minutes in a fitness app and still not complete a workout. A user may open a finance platform several times and still delay a decision. Someone on a social app may swipe through dozens of profiles and leave with no meaningful connection, no meeting arranged, and no clearer sense of what they are actually looking for. In each case, the platform can still record engagement, even while the user experiences indecision, overload, or disappointment.

That is why the intention-action gap has become such an important issue in consumer technology. Most people do not fail to act because they lack interest. They fail because friction builds up. Too many options, poor timing, and repetitive interfaces make follow-through harder than it should be. Traditional engagement design often worsens that problem because it rewards prolonged activity instead of successful resolution.

How behavioural AI changes the model

Behavioural AI is valuable because it looks beyond isolated clicks and interprets patterns in context. It can identify hesitation, momentum, preference shifts, and likely drop-off points. More importantly, it can respond to those signals in ways that make decisions easier and outcomes more achievable.

That changes the app’s role. Instead of acting primarily as a feed, a storefront, or a passive interface, it starts to function more like an active guide. It can narrow choices when users are overwhelmed, surface the next best action when intent is clear, and adapt when behaviour suggests a mismatch. This can mean recommending fewer but better options, improving prompts, changing timing, refining compatibility logic, or reducing unnecessary steps between interest and action.

The commercial relevance of this shift is growing. SAP reported that 82% of UAE marketers say AI is central to their personalization efforts, yet only 31% of consumers believe brands actually personalize content to their needs. Data and automation alone are not enough. Relevance depends on using insight in ways that feel useful, proportionate, and credible to the user.

From digital engagement to real-world outcomes

Behavioural AI becomes especially powerful in categories tied to everyday behaviour and human relationships. In social discovery, for example, the challenge has never been a lack of available profiles. It has been helping people move from superficial activity to meaningful connection.

That is where a social platform like MAXION sits within a more important conversation about the future of consumer apps. Success should not be measured only by how many profiles a person sees or how long they stay active on the app. It should be measured by whether the app improves the quality of interactions and increases the likelihood of real-world meetings.

Behavioural AI can support that by learning from interaction patterns. It can identify where conversations stall, what kinds of introductions lead to better follow-through, how timing affects responsiveness, and which recommendation patterns create genuine alignment rather than short-lived engagement. That creates the possibility of designing around success signals that matter outside the app.

This is also highly relevant in the UAE, where AI adoption is already part of everyday life. KPMG reported that 97% of UAE respondents use AI for work, study, or personal purposes. That level of familiarity creates a more sophisticated user base.

The broader point is that consumer AI is becoming more outcome-oriented. Whether the category is education, wellness, finance, or social connection, the products that stand out will be those that reduce noise, respect user intent, and drive real-world progress. The next generation of successful apps will be defined by how effectively they help people do something worthwhile with them.

Alexander Merkushev, Head of AI projects, Yango Tech

Business pressure rarely arrives in a neat or predictable form. It builds through overlapping demands, such as customers expect faster responses, regulators expect tighter control, leadership teams need clearer visibility, and frontline staff are asked to deliver all of this through systems that often do not move at the same speed. In stable conditions, organisations can usually work around those gaps. Teams compensate manually, service holds together, and inefficiencies stay partly hidden. In high-pressure environments, that buffer disappears. Slow workflows, fragmented systems, and manual bottlenecks become visible very quickly because the organisation no longer has the time or flexibility to absorb them. That is where the case for AI agents becomes much more practical. AI agents are most valuable when they allow businesses to extend operational capacity, where adding more people alone does not solve the problem fast enough.

This is especially relevant in the UAE, where digital maturity has raised expectations across both public and private sectors, with the UAE ranking 11th globally in the UN’s 2024 E-Government Development Index. This stronger digital environment has also raised expectations. Businesses need tools that can help them move quickly, stay consistent, and maintain control when pressure rises.

From Tools to Agents

With around 84% of GCC organisations adopting AI, it must prove its operational value. This is where autonomous AI agents stand apart from basic assistants. The lesson from digital transformation and automation is that technology creates the greatest impact where work cannot be carried out reliably at scale by people alone. That usually means high-volume, repetitive, rules-based, or time-sensitive tasks that still require consistency and traceability. A conventional assistant can answer a question, retrieve a document, or draft a message. An AI agent can operate across workflows, connect with enterprise applications and data sources, retrieve the information needed for a task, trigger an action, and escalate the case when human judgment is required. AI agents are less like a front-end convenience and more like a digital workforce layer that supports execution inside the business.

Keeping Service on Track

Customer service is often the first area where this becomes visible because it sits at the intersection of urgency, expectation, and reputation. When volumes rise, even strong teams can be slowed by manual routing, repeated verification, inconsistent answers, or language limitations. A customer support agent can handle thousands of routine queries across languages and channels without making customers wait for basic answers.

In fact, enterprise deployment data points to AI agents that can operate in 70+ languages, integrate with core business platforms such as CRM and support systems, and scale to handle 100,000+ interactions per day. Outcomes include 95% first-contact resolution, a 70% reduction in calls, and around 40% lower support costs. In a high-pressure environment, the benefit of an AI agent is that it helps the organisation respond at scale without allowing service quality to collapse under volume.

Compliance Under Pressure

Businesses often wrongly assume AI will automatically make operations faster, but the speed needs to be usable inside a controlled environment. If an agent cannot follow policy, log its actions, flag discrepancies, and escalate exceptions correctly, then it simply moves the risk somewhere harder to see. Well-designed AI agents can reduce delay by supporting documentation checks, rule-based workflows, anomaly flagging, and routing complex issues to the right human decision-maker while maintaining auditability.

For instance, Yango Tech’s AI debt collector agent can support repayment workflows, structure payment plan discussions, apply pre-set compliance rules, and manage routine follow-ups while flagging exception cases. A document analysis agent can review procurement files, compare them against required fields, and flag inconsistencies. The limits of disconnected tools are exposed very quickly in high-pressure environments, and businesses need systems that can work inside the operational environment that already exists.

Why digital workers are becoming relevant

In volatile conditions, where teams are stretched, leaders do not benefit from more dashboards or longer reports. Current industry findings show that organisations can lose 30 to 50% of efficiency to repetitive tasks. Too many skilled employees still spend time gathering updates, moving information between systems, or preparing routine reports instead of focusing on judgment, service recovery, and problem-solving. AI agents can absorb that repetitive load and help teams concentrate on higher-value work. They can surface relevant data from multiple systems, summarize key trends, identify pressure points, and reduce the delay between an operational change and a management response. Their role is to help leadership reach judgment faster, with better operational visibility and less reporting friction.

High-pressure environments reveal which technologies can support real execution. AI agents are most useful where organisations need to operate at a scale, speed, and consistency that people alone cannot sustain manually. But that only works when the system is designed with the right guardrails. Service quality, oversight, escalation logic, and traceability cannot be added later as an afterthought. Companies like Yango Tech create production-ready AI agents for high-pressure and fault-sensitive environments and help organisations deploy them in a governed, resilient, and reliable way under real operational strain.