Hi , On-Balance Issue 183: Taming AI needs DEI

Hello there and welcome back to On Balance, my fortnightly round-up of news, views and research around the issue of diversity, equity and inclusion.

Taming AI needs DEI

For much of 2025, it has seemed as though leaders have taken potshots at DEI with full impunity. Businesses have walked back public commitments to inclusion and initiatives aimed at improving equity, repudiating them loudly, with little apparent concern about the ultimate impact of their words and actions.

A hostile climate is ultimately uncomfortable for everyone who has to live in it.

And so it’s unsurprising to see the signs that the chickens may be coming home to roost in recent weeks for those who have surfed the Trumpian wave through corporate backtracking on inclusion or anti-DEI activism.

At Target, CEO Brian Cornell recently said goodbye to the business after the US retailer’s DEI rollbacks of early 2025 translated into falling sales after a customer boycott. This week saw reports of an exodus of top talent at Elon Musk's businesses due to disillusionment with his political manoeuvring and frustration with his leadership style, cited among the factors for departure.

Losing customers and falling share prices are uncomfortable for any business - but for a good leader, they should be fixable.

But when we connect the DEI rollbacks to the other big story of 2025 - the mass adoption of AI and the preparation for agentic AI in business - the leaders who have abandoned their commitments to equity and talent could face a far bigger reckoning, with an impact that lasts far beyond the discomfort of a quarterly earnings call.

In the past year, we have learnt much about what AI will offer to the workplace. In automating, working faster, and saving time on tasks that otherwise require significant human effort and knowledge, it is delivering as promised. But we cannot trust these tools to operate without bias or error. In recruitment, it has downgraded women’s CVs. In healthcare, it has underdiagnosed minorities. It regularly hallucinates facts, references, or calculations. We understand that AI reflects the strengths - and the weaknesses - of the culture that created it. Which is why every output must be tested, not assumed.

This fundamentally changes what is required of our leaders today and in the future. It is no longer just about making decisions; it is about judging whether AI's decisions can be trusted.

That demands habits inclusive leadership develops: spotting blind spots, questioning assumptions, listening to dissent, weighing competing arguments, pausing before signing off. Diverse teams provide this capability. They challenge groupthink, surface edge cases, and ask, "What have we missed?".

Leaders who build inclusive teams are more likely to listen than talk, to move beyond averages to understand outliers, to demonstrate the humility that prevents catastrophic errors. These habits matter everywhere - in financial modelling, supply chain planning, and reputation management - but they matter most when deploying technology that amplifies bias at scale.

Rolling back DEI doesn't just weaken your connection to customers or deprive you of diverse experience. It strips away the scaffolding that makes your organisation ready for AI.

AI will amplify whatever leadership brings to it. Inclusive, humble, questioning leaders will use it to make their organisations smarter and more resilient. Narrow, short-term leaders risk hard-wiring bias and accelerating decline.

It's true that the extreme cases make the headlines: Target and Musk are outliers. But headlines can make the weather, and the current climate has made many organisations quieter about their commitment to inclusion, equality, and diversity - even though, from what we see, the work continues.

So while the prevailing narrative may be uncomfortable, those organisations getting on with developing inclusive leadership capability, quietly and consistently, should feel confident. They are embedding exactly what they'll need for the disruption ahead. While the rollbacks grab attention, it's those who held the line, without fanfare and despite the noise, who will have the competitive advantage precisely because they maintained the capabilities AI governance requires.

So the real story isn't about the handful who walk. It's about whether your organisation is holding firm and building the judgment, humility, and diverse perspectives that AI demands. That work matters more now than it ever has, and the reward is best measured in commercial success rather than headlines.