The Future of Work Isn't Waiting: What We Heard and Insights from UNLEASH Paris

Our team spent two days at UNLEASH Paris listening to Fortune 500 leaders, pioneering HR practitioners, and forward-thinking organisations share what's really happening as AI reshapes talent acquisition. Here's what stood out for us across the event - and what it means for anyone building the workforce of tomorrow.

‍

If you attended UNLEASH Paris this year, you felt it. If you watched from afar, you probably sensed it.

AI wasn't just a topic at this year's event - it was the driving force.

Every session. Every vendor stand. Every conversation. The question wasn't whether AI would transform how we work, but how quickly organisations could adapt without losing what makes them human.

Our team left Paris with pages of notes, dozens of conversations replaying in our heads, and one overarching observation: we're at a real inflection point. The organisations that will thrive aren't racing to deploy AI fastest - they're rethinking their foundations entirely. Fixing processes that don't work. Building data architectures that actually deliver value. Creating cultures where people and technology amplify each other, not compete.

Here's what we learned.

When Everyone Has AI, Culture Becomes Your Competitive Edge

Bruce Daisley opened the conference with a provocative premise: when "intelligence on tap" becomes universal, competitive advantage shifts dramatically. If every company can access the same AI capabilities, what actually sets you apart?

His answer, grounded in occupational psychology research: the human layer.

The data he shared painted a picture many of us recognise. Under-40s are showing significant declines in conscientiousness and increases in neuroticism. Extroversion is down. We're spending 99 minutes less per day with other people than we did pre-pandemic. Work has become more fragmented, more isolated, more transactional.

The differentiator for organisations is the layer of culture, Daisley argued, and leaning into and celebrating the autonomy of workers, their identity, and the sense that we're all in it together.

He broke down what actually drives motivation and resilience:

• Autonomy (the biggest motivator)
• Identity (respect, belonging, having a best friend at work)
• Mattering (the belief that you're significant to others)
• Connection ("we're all in this together")

From a skills science perspective, this reinforces something we've long believed: technical capability alone doesn't predict success. The organisations investing in understanding cultural fit, team dynamics, and behavioural skills alongside technical competencies are building something AI can't replicate - environments where people actually want to show up and do their best work.

Sharon Doherty, Chief People and Places Officer at Lloyds Banking Group, reinforced this point, saying "If you're in human capital, you are the vibe king or queen - the company will follow your lead."

You Can't Layer AI on Top of Broken Processes

This was the refrain we heard in nearly every practitioner session, and it's worth repeating because the temptation to do exactly this is enormous.

Leaders from HelloFresh, Airbus, British Airways, and Tesco all emphasised the same principle: AI doesn't fix bad processes - it exposes them.

Syed Ali Abbas, VP of People at HelloFresh put it plainly: "The big trend here is not AI, it's the datafication of everything. We're changing data architecture, quality, flow, governance. AI takes data and gives outcomes faster and better. We need to prepare for this."

Think about what that means. Before AI can deliver value, you need:

• Clean, structured data
• Clear processes that actually work
• Governance frameworks that make sense
• An understanding of what problems you're actually trying to solve

Eva Linares, Head of HRIS and HR Data at Transdev drove this home when discussing vendor selection: "Technology isn't the silver bullet. Think about your why, your processes and ecosystem around it. Once you've figured that out, technology works." She added a warning that should make every procurement team pause: "When vendors come to you, half of what they're going to sell doesn't exist. Be ready to challenge what they're telling you - dig into it, understand limitations. 80% of AI projects fail - ask yourself why."

The skills science angle here is critical. Many organisations are discovering that AI can infer skills from CVs or past experience, but without validated assessment science, you're making expensive decisions on incomplete data. Skills science provides that foundational universal data layer when it comes to talent - measurable capabilities that AI can actually work with. It's the difference between asking AI to guess what someone can do - or worse, draw on generic examples, versus giving it accurate context and inputs to optimise against.

The Stakeholder Challenge: 80% of Your Time Will Be Building Buy-In

Here's a stat that surprised us, even though perhaps it shouldn't have: Eva from Transdev shared that 80% of her project time is spent talking to operations, building networks of key people across different countries and functions.

Not building the solution. Not implementing the technology. Building buy-in.

The panel on "Securing Board Buy-In" was one of the most practical of the conference. Josephina Smith (British Airways), Eva (Transdev), and Anna Carlsson (HR Digi) shared hard-won wisdom:

How to prepare for success:

• Know the macro environment (senior stakeholders already do - you need to as well)
• Understand your organisation's specific context (what they call "contextual intelligence")
• Remember that boards care about three things: make money, save money, keep money
• Articulate value creation simply - not with 100 slides
• Link to real business indicators, not just HR metrics

Eva's approach: "Imagine yourself as a startup. My project is my startup, and I have to sell to my investors, my board. Also, think about the people I touch with my project - managers, employees, functional leads - what am I doing for these people? Ask them if they see the value. If they don't, it'll be tough to drive change."

She also shared a brilliant metaphor: "It's important to have people disagree with what you believe. When you ski, you need to fall - if you don't, you're not trying hard enough."

Cecilia Sandberg, SVP CHRO from Atlas Copco reinforced this, sharing that 90% of AI work is stakeholder involvement and management. Their approach to getting 24 divisional presidents on board? Start outside-in -what are the real pain points? Reduce time to hire, improve diversity and quality, improve internal mobility, reduce time to competence. Business challenges can form your use case. Stakeholders are data and fact-driven. Gather together, talk it through.

The takeaway: transformation isn't a technology problem. It's a people problem that technology can help solve - if you've done the groundwork.

Transformation Is a Non-Negotiable

Louis Bosche, Head of HR Tech at Pernod Ricard shared impressive numbers around their skills-based transformation. Skills are now at the core of how they "build, buy, and borrow" talent.

But the real insights came from the less polished moments - the conversations about what it actually takes to get there.

Lloyds Banking Group is two decades into their transformation journey. Sharon Doherty described it as wearing "three important hats":

1. Storyteller - Tell a story across decades - past, present, future. Respecting history is vital. The language you use creates a foundation when navigating change. Don't erase where you've been; build the bridge to where you're going.
2. Tough lover - Lean into the tough legacy things early. The things your senior leadership team probably knows about but finds too hard to change. Sharon gave a concrete example: they removed compressed working weeks, engagement scores dropped, but then they established better flexible working practices - and by 2025 saw huge increases in positive employee engagement results. "Face into tough change really early - send a flare up."
3. Disruptor - "Yes, we're disrupting, but disruption starts with ourselves. We all need to go back to school - AI, tech, agile as well - we have to learn so we can play a big part and lead an exciting future."

The takeaway: transformation at this scale isn't about deploying tools - it's about having the conviction to lead through discomfort, the wisdom to honour what came before, and the courage to challenge what no longer serves you.

The "Cognitive Debt" Concern: Are We Building Capability or Dependency?

One of the more sobering themes was around early careers and what happens when we over-rely on AI without building foundational capabilities.

HelloFresh's Syed called it "cognitive debt" - the risk that people using AI extensively early in their careers aren't building the neural pathways required for reasoning, planning, and problem-solving. "We might end up with a lower quality workforce unable to handle cognitive load," they warned.

Vincent Dupuis, VP HR Digital and AI at Airbus, framed it even in important context. As a company producing aircraft where lives depend on their products, they ask what if frontline workers rely completely on AI, and if something goes wrong, they don't remember how to fix it? Are we creating more risk than opportunity? They don't want to create what could be "brittle knowledge - where the tree falls because the roots are too thin."

The philosophy: augment everyone, automate low-value tasks, but maintain critical thinking for critical tasks. "Technology will enable everything, but what do you want as a company to become? Think not in tech but what the company is about."

Paddy Hull, Global Head of Talent, Culture, and Leadership from Heineken shared a personal example that illustrated the shifting trust dynamics. He asks Microsoft Co-pilot weekly to give him feedback as a "trusted advisor who wants the best but wants to give honest feedback about how I'm turning up. Results have been astonishing - I've got feedback I never had from human managers that was very helpful. Actionable advice on steps to take."

The question for organisations: how are we designing AI to 10x human capability and experience, whilst also ensuring we don't erode it?

The Reality Check: What's Actually Working (and What Isn't)

The panel on "Agentic AI in HR" brought together Victoria Murphy (Head of HR Integrated CoE Operations at JLL), Kris Legroe (Director People Technology at VMO2), and Chris Browne (Head of Product - Talent Acquisition at Tesco) to discuss what's working on the ground.

The operating model question: Should AI initiatives be HR-led or IT-led?

Kris from VMO2 argued it should be HR-led, focusing on operational efficiencies and identifying use cases where AI creates real value in talent and learning workflows. Chris from Tesco agreed on HR leadership but emphasised the critical need for tech enablement—too many impressive proof-of-concepts fall apart in production because they lack understanding of actual business processes. Victoria from JLL highlighted the prioritisation challenge—with so many vendors offering similar capabilities, they focus on decisions within their ecosystem (tech, procurement, HR, legal) based on resource availability and ability to show impact quickly. Her approach: patches rather than global rollouts to demonstrate success before scaling.

The orchestration challenge became a common theme. Organisations learned quickly that deploying AI without proper orchestration leads to confused outputs and context-switching errors. The winning approach: specialised agents with clear orchestration layers rather than single agents trying to handle everything. Breaking down functions improved both quality and cost management.

Their advice:

• Kris: Start with simple, small use cases
• Chris: Think of agents not as people but as tasks that need to be done, skills that augment the workforce
• Victoria: Deconstruct. Don't put bots over bad processes. Clean it up, and have fun with bots.

Accenture shared their "zero touch hiring" pilot - moving from 7 million applications to 120,000 hires. In one location, they replaced 4,000 skills interviews for Java with digital assessment, saving 4,000 hours of potential billable work every week. Candidate experience and timelines improved from a month to a week or two.

But Tom Sayer from Accenture raised a critical challenge: legal defensibility. While Accenture invested in full job analysis and incumbent studies to ensure their process was defensible, the timeline was lengthy. As organisations move faster with agentic AI, the industry faces a tension - maintaining legal rigour whilst accelerating implementation. The current job analysis approach doesn't scale at the pace AI enables.

This is precisely where validated skills science becomes invaluable. The organisations that will move fastest are those with robust assessment frameworks already in place - validated measures that can be adapted quickly rather than built from scratch for each use case. It's one of the reasons we've invested heavily in our AI-powered role skills profile builder: helping organisations establish that defensible, scientific foundation faster, so they can move at the pace AI enables without sacrificing accuracy and rigour.

HR's Moment to Lead (If We're Willing to Grab It)

Peter Lynch, CPO at Cardinal Group Companies, delivered what might have been the conference's rallying cry: "The future of organisations is ready for someone to grab it - nobody is better suited than heads of HR."

He called this the "capture the flag moment."

His premise: the merge of HR and IT creates both dilemma and opportunity. The edge in the future of work is human + AI, culture + innovation, leadership + tech.

Now is the time were HR doesn't just have a seat at the table, it's their time to own the table.

At Cardinal Group, they're seeing 73.6% monthly active users and 27.1% daily active users of AI tools. But Lynch's focus wasn't on the technology - it was on what he called "the power of messy humanity" and "the UGLY advantage." UGLY stands for unique (not how your make others see you, how you make them see themselves), Gutsy (it's not about the start, but the ability to recover from failure), Likeable (conversations based on what people need, not what they want), and Yoked (the power of trust in the whole being greater than the sum of the parts).

He shared a story about Belgian horses. Farmers measured the output of two horses working together, and the results didn't match the assumption - it was even better. The magic of culture - one team "yoked together" - created exponential value.

"Companies have to be able to access that magic," Lynch argued. "It happens through CHROs."

Sharon Doherty from Lloyds echoed this: "If HR doesn't shape the future of work, who will? You have to shape, design the workplace, recognise teams versus individuals. It needs to be shaped."

‍

What This Means for How We Think About Talent

Walking out of UNLEASH, a few things became crystal clear:

1. The market is moving faster than many anticipated. AI adoption in talent has gone from experimental to essential. Organisations are moving - implementing agentic AI, building skills catalogues, rethinking entire workflows - without waiting for perfect solutions. They're piloting, learning, and scaling whilst others are still debating.

2. There's a massive awareness gap. Gallup reported that 93% of Fortune 500 CHROs say they're deploying AI, but only 33% of employees know about it. McKinsey noted that 80% of companies are using AI, yet only 20% are creating measurable value. The implementation challenge is real.

3. Culture isn't a soft skill - it's the differentiator. When technology becomes universal, how well humans work together becomes the sustainable competitive advantage.

4. Data quality is the foundation that determines whether AI helps or hurts. The "datafication of everything" creates opportunity only if that data is structured, validated, and meaningful. For talent and people decisions, skills science is becoming the essential data layer - the structured, validated measurement of human capability that AI needs to deliver real value rather than expensive guesswork.

5. Stakeholder management is 80% of the work. Technology implementation is the easy part. Building buy-in, demonstrating value, and bringing people along - that's where transformation actually happens.

6. HR has an unprecedented opportunity to lead. But only if we're willing to disrupt ourselves first, develop AI literacy, and become architects of the future of work rather than administrators of legacy systems.

From a skills science perspective, what excited us most was seeing organisations realise that AI without valid, fair, transparent decision-making frameworks is just expensive automation. When you combine AI with proper skills science - validated assessment, robust methodology, and a clear understanding of what actually predicts performance - that's when the magic happens. Substantial cost savings, faster hiring, and better quality decisions.

Key Takeaways

If you only remember three things from UNLEASH Paris:

→ Culture is your competitive advantage when AI becomes universal. Invest in the human layer—autonomy, identity, connection, and psychological safety.

→ You can't bolt AI onto broken processes. Fix your foundations first: data architecture, clear workflows, validated measurement frameworks.

→ Transformation is 80% stakeholder management, 20% technology. Build your internal network before you need it. Make the business case simple. Link to real outcomes, not HR metrics.

The future of work is being written right now. Make sure you're holding the pen.

The Conversations Continue

UNLEASH Paris made one thing abundantly clear: we're not in a transformation phase anymore. We're in reinvention.

The questions aren't "should we implement AI?" but "how do we implement it at our organisation's pace whilst maintaining what makes us human?"

The answer won't be the same for everyone. Scale matters. Industry matters. Culture matters. Starting point matters.

But the principles are consistent: start with clear problems, build on solid data foundations, bring stakeholders along, and never lose sight of the human element that makes great teams perform.

We left Paris with more questions than answers - which feels exactly right. The organisations that will succeed aren't the ones with all the answers today. They're the ones curious enough to keep asking better questions.

Want to dig deeper into any of these themes? Curious how skills science can provide the data foundation for your AI initiatives? We'd love to continue the conversation.

The Spotted Zebra team attended UNLEASH Paris in October 2025. All quotes and attributions in this article are from sessions and conversations at the event. Our thanks to the practitioners and thought leaders who shared their experiences so openly.

‍

‍

‍