Lead The Workforce Or Lose Your AI Transformation

“We were talking about you, not to you” was a common quip growing up, used when someone turned around expectantly after overhearing their name mentioned.

Similarly, the current AI discourse often cites the 3.5 billion workers who are beginning to be disrupted, but so far, it is talking about them, not to them.¹

To date, frontier labs have primarily engaged investors using big existential messages about AGI, human replacement, and an AI utopia. This approach has worked well for them, and on March 31, 2026, OpenAI closed a record-breaking funding round, with $122 billion in committed capital at a post-money valuation of $852 billion.

There has also been considerable “thought leadership” directed at executives and shareholders from the broader AI ecosystem,² pitching dramatic productivity gains and an AI-first redesign – now with an agentic slant.

However, I could not find a single substantive communications campaign aimed at workers about how AI will enhance their future. When I asked ChatGPT to search, it responded that the lack of examples was “stark”. Even the few pieces that appeared to focus on the benefits of tools for employees were only talking to their organisations.

The missing alignment

Some bold architects of the AI revolution will say worker discomfort is a transitional inconvenience, soon to be resolved by their replacement, so communications are superfluous.

However, I have been involved in technology-driven transformations for more than three decades, and old-fashioned human buy-in has been a critical factor in every one of them.

A growing body of research finds similarly that AI implementation failures are driven less by technical limitations and more by human factors, including organisational misalignment and weak change management.

A systemic lack of alignment on AI is already evident in a string of recent reports on worker attitudes to its adoption. The latest Stanford HAI AI Index report shows a clear and widening gap between optimism at the top (experts, executives) and scepticism among the public. Only 33% of American workers expect AI to improve their jobs, and 52% are nervous about its impact.

This is not surprising. Even though the AI ecosystem does not speak directly to workers, they are comprehensively exposed to narratives about themselves. It is not difficult to know what many investors and executives intend to do.

The missing motivation

However, despite all the talk, AI has not yet displaced jobs in a measurable way at a macroeconomic level. Yes, some companies are slowing hiring or even making layoffs, but typically based on the prospect of AI productivity rather than its current reality.³

The net result is a workforce on which employers are still entirely dependent, feeling threatened and anxious about AI.

In these conditions, how many workers will enthusiastically train an agentic platform to do their job if there is no path forward that safeguards their future? Who will want to coexist benignly with the technology that replaced their friends? A Writer survey even found that 29% of employees admitted to sabotaging their company’s AI rollout, rising to 44% among Gen Z.

Please read that last extraordinary finding again.

The missing leadership

To be clear, I am not arguing for guaranteed employment or resistance to AI-driven change. My call is for meaningful worker engagement, for creativity, and for a mindset that rejects the idea that human and AI work are in a zero-sum relationship. This is, in other words, a leadership challenge – and an urgent one.

Here is some action we can take in the short term:

1. Engage with staff about AI to find out what they really think and feel, and how beneficial current usage is on the ground. Capture their ideas. Go and talk to some of them yourself
2. Invest in AI training for everyone; either they become more informed employees in your organisation, or have better prospects outside if they are displaced
3. Adopt a human-centric design approach that focuses on work that no one wants to do and avoids cognitive atrophy. This is easy to say, but not proving easy to do
4. Think creatively about how displaced workers can be retrained, redeployed or at least replaced with other humans for growth rather than just efficiency. This may not be possible, but it should always be considered
5. Make the distribution of AI-driven productivity gains explicit. If value is created, define how it will be shared with employees, not just shareholders
6. Establish a work environment in which human and artificial intelligences coexist in synergistic ways – making the best of both sets of attributes and removing a sense of contention

The missing returns?

Just as we enter a new phase of capable, agentic AI that unlocks real productivity, we risk repeating the mistakes of earlier technological transitions by deprioritising engagement with human workers; only this time, the effects of our errors could be on an epic scale.

The corrective action is AI leadership: engaging, aligning, and motivating staff towards an ethical but commercially productive outcome.

It is time to start talking to, not just about, workers in the AI discourse.

I believe that returns on more than $2.5tn of planned investment in 2026 alone depend not on technical advancement but on this very human leadership imperative.

Footnotes

1. I recognise this is also true of The Hornet – but I have sought to broaden my audience through open-access community lectures. No doubt, I have more to do.

2. Including AI labs, specialist media, startups, and consultants.

3. I have become obsessed with parsing public statements, looking for this distinction.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.