Issue 002: The AI reckoning nobody in HR is having

Most organisations have an AI strategy. Almost none of them have thought seriously about what that strategy is doing to their people, to their confidence, their sense of security, their trust in leadership.

The companies winning the AI transition right now are not the ones with the best tools. They are the ones where leadership sat down and had an honest conversation with their workforce about what was coming, what it meant, and what they were going to do together. That conversation is rarer than it should be.

This week is about the leaders who had it, and the templates you can use to have it yourself.

Welcome to Issue 002 of The Work Life Reporter. This week you get:

Let’s get into it.

After Unilever announced significant AI investment, internal surveys showed 71% of employees were anxious about job security. Rather than wait for rumours to fill the silence, the CHRO partnered with the CEO to issue a direct, plain-English communication to all staff, naming specifically which roles would be augmented by AI, which would change significantly, and which the company was actively protecting.

The move worked because it traded vagueness for honesty. Ambiguity is not kindness. When people don't know what's coming, they assume the worst and start looking elsewhere. Unilever understood that the cost of a difficult conversation is always lower than the cost of lost talent. Clarity on AI isn't just ethical, it's a retention play.

In early 2024, Klarna made headlines announcing AI would replace 700 jobs. What didn't make the same headlines: the weeks of internal chaos, manager resignations, and crisis management that followed. Behind the scenes, the People team ran urgent "listening circles" — not to walk back the announcement, but to understand what people were actually afraid of.

What they found changed everything. The fear wasn't purely about job loss. It was about dignity, people needed to know they mattered regardless of what a machine could do. The circles became a model: small groups, no cameras, no notes taken by management. When you give people a genuine place to speak, they usually say something worth hearing. The format spread across the HR function because managers started asking for it.

When Microsoft started embedding Copilot across the organisation, its People Analytics team built an internal tool — the AI Readiness Index — that assessed not just technical readiness but behavioural and cultural readiness at team level. Managers got a simple dashboard showing how their team scored on confidence, understanding, and adoption. The scores weren't used in performance reviews. They were used to trigger support.

The finding that changed their approach: teams with the lowest adoption weren't the least technically capable. They were the ones whose managers had never talked to them about AI at all. Technology adoption follows psychological safety, not access.

When HubSpot launched its AI skills programme, most companies framed theirs as "become more productive." HubSpot took a different route. Their CPO positioned AI literacy as a career accelerator: employees who completed the AI skills pathway got priority access to internal mobility opportunities, stretch projects, and quarterly mentoring sessions with senior leaders.

The result was a 34% higher voluntary completion rate compared to a traditional compliance-framed rollout. The framing was the product. People don't do things because you tell them it's good for the business, they do them when they can see a clear connection to something that matters to them personally. HubSpot's People team understood that motivation lives in career narrative, not duty.

When Vodafone began its AI-led transformation in 2024, HR Business Partners were trained on a specific protocol for when employees raised job security concerns in one-to-one conversations. Rather than deflecting with "we're not at that stage yet" or over-reassuring with "your job is safe," HRBPs used a structured three-part response framework the People team had developed internally.

Part one: acknowledge the concern directly. Part two: share honest context, what the organisation knew and didn't, without speculation. Part three: give agency, what the employee could actually do right now, decisions they had real control over. The framework spread because managers started asking for it too. Most managers want to have these conversations well. They just don't have the language. This gave them it.

Most organisations have written an AI policy in the last 12 months. Most employees have no idea what it says. The gap between policy and practice in AI adoption isn't a communications failure, it's a design failure. Policies written by legal and IT for compliance rarely reach the people they're meant to guide. They're long, passive, and written for the edge case rather than the everyday.

The fix isn't a better document. It's moving the policy off the intranet and into the workflow. The organisations seeing actual behavioural change are the ones embedding AI guidance at the moment of use: a one-pager pinned inside the AI tool itself, a two-minute manager-led briefing at the start of the week, a checklist before anything AI-generated leaves the building.

If your AI policy lives in a SharePoint folder no one opens, it's not really a policy. It's liability cover. The question worth asking: if a team member used AI irresponsibly tomorrow, would they know they'd crossed a line, or would they find out when something went wrong?

There is a culture risk in AI adoption that gets almost no airtime: the risk of uneven adoption creating a two-tier workforce. In most organisations right now, you have a group of people who are actively using AI, getting faster and more capable, building confidence, and a group who aren't. Either because they're afraid, under-resourced, or have a manager who's never mentioned it. Over time, that gap becomes a performance gap. Then a career gap.

People leaders need to track AI adoption the way they track engagement. Who's using what? Which managers are enabling it and which aren't? What does the adoption picture look like across demographic lines, age, tenure, department, level? Because if the pattern mirrors existing inequities in your organisation, AI will amplify them, not erase them.

The organisations doing this well aren't measuring e-learning completion rates. They're asking managers directly: who on your team could really benefit from support here? And then following that question with resources, not rhetoric.

"We're committed to responsible AI" appears on the website of almost every company deploying AI right now. What it means in practice varies enormously, from robust governance frameworks and ethics oversight to, essentially, nothing. The difference between the two is not intent. It's infrastructure.

Responsible AI at the people level means three things: people know what data is being used to make decisions about them, people have a genuine way to question those decisions, and the humans in the chain haven't stepped back just because the algorithm gave them cover. The third one matters most in HR. When an AI tool flags an employee for performance risk, what happens next? Is a human making the call, or just signing off on it?

Culture leaders have a specific role here that doesn't get named often enough: building the management muscle to use AI tools as input, not verdict. That requires training, and it requires holding managers accountable when they don't use that judgement. Governance without manager enablement is compliance theatre.

Four months ago, Alisa Latty-Alleyne launched HebrewsEleven, an ethical AI advisory consultancy supporting organisations, boards, and mission-led institutions to adopt AI responsibly, without losing what matters most to them. She brings 20 years of senior leadership experience across corporate and the charity sector, and is currently completing an Executive MBA in AI and Digital Transformation at Warwick Business School, where she was awarded the Inspiring Female Leaders Scholarship.

She is also building H11 Labs, the social impact arm of HebrewsEleven, specifically focused on ensuring AI is accessible and equitable for underrepresented communities, faith groups, and young people who risk being left behind as AI advantages compound across better-resourced organisations.

In June 2025, she received the She Leads Woman of Distinction Award, presented by HRH Sophie, Duchess of Edinburgh.

In this week's Spotlight, we sit down with Alisa to unpack what responsible AI leadership actually looks like in practice, why most organisations are getting the people side of AI wrong, and what it means to lead with confidence when the ground keeps shifting.

Alisa's starting point is that most organisations are treating AI adoption as an IT project when it is actually a cultural transformation. The technology will keep evolving regardless of what any one organisation decides. What won't evolve on its own is how leaders communicate, make decisions, and show up for their people. The organisations doing this well have senior leaders who've done the personal work to understand what they don't know, and who are honest about it rather than performing confidence they don't have.

In her advisory work, Alisa finds consistently that employee anxiety around AI is less about the technology itself and more about the absence of honest conversation from leadership. People can handle difficult news. What they struggle with is the vacuum that forms when leadership goes quiet. The first question she asks every organisation: when did your CEO last speak to your entire workforce about what AI means for them, specifically, not in abstract terms?

Through H11 Labs, Alisa is building programmes for communities who risk being left behind as AI advantages compound across better-resourced organisations. For people leaders, this has a direct organisational parallel: if you're not actively tracking who in your workforce is benefiting from AI adoption and who isn't, you're not doing responsible AI. You're just doing AI.

Alisa's view on governance is direct: most AI policies are written to protect the organisation, not to guide the people inside it. The gap lives in the middle layer, managers who haven't been briefed, trained, or given the language to have these conversations with their teams. Governance without manager enablement is compliance theatre. Closing that gap is the people function's job, not legal's.

One of Alisa's core speaking topics is "AI Without Fear", and her point isn't that leaders should be fearless. It's that fear managed privately becomes paralysis, while fear acknowledged openly becomes a starting point. The leaders she most admires right now are the ones who say: here's what I know, here's what I don't, and here's how we're going to figure this out together. That's not weakness. That's the job.

Thank you, Alisa!

Connect with Alisa on LinkedIn or find out more at hebrewseleven.com

And that's a wrap for Issue 002. Something land? Hit reply and tell us which section you're taking back to your team.

Follow The Work Life Reporter on LinkedIn for what we share between issues.

PS. Know someone who leads people and would find this useful? Forward it on. Ten seconds to share, a lot longer to write.

PPS. Got a story, a leader, or a topic you'd like us to dig into? Reply with your ideas.