If one wanted to summarise the articles we've written about our DEI strategy over the last years, it would be this: "We are continuing to make progress, but we will never be finished." The more we've investigated where AI development is heading, the more this feels like an understatement.
Roughly 70 nationalities work under the Netlight umbrella, and around 40% of our employees are women. Neither number is incidental. But how do those numbers hold when you look at professionals focused on data and AI specifically? Yet take gender alone: according to the Stanford AI Index 2026, the share of women among AI authors and inventors has been essentially flat since 2010; across 21 countries tracked, no country comes close to parity. Fifteen years of AI becoming one of the most consequential technologies in human history, and the gender balance of the people building it hasn't moved.
Even by market standards, our numbers may look better, but we are still struggling to find female talent in data and AI. And even when we do hire diverse talent, that doesn't automatically translate into diverse teams on the ground. A delivery team where one person with a different background is outnumbered ten to one has placed a disproportionate burden on a single person, not built a diverse team.
One of our consultants was the only woman on a team of roughly a dozen, the only person with her background. She pointed out blind spots, asked uncomfortable questions, proposed solutions. Her points were dismissed. Only to be adopted months later after costly mistake that were identified in the “mandatory” validation phase. Whether that happened because she was seen as an outsider is impossible to prove. What is clear is the outcome: a costly delay, a preventable mistake, and a solution that was always in the room.
However, focusing only on the risks diverse team help to mitigate is missing the bigger story. The real value is not that diverse teams prevent bad decisions. It is that they enable better ones. The most successful AI solutions are rarely the result of better technology alone. They are the result of better decisions, shaped by teams that challenge assumptions, bring different perspectives, and understand a broader range of users and real-world situations. Diversity is a competitive advantage because it helps teams make better decisions, build better AI, and ultimately achieve better outcomes.
In consulting, that matters. Clients approach us to go beyond risk reduction and assessment. They come to us to make better decisions, build better products and create better outcomes. The quality of those outcomes is directly influenced by the perspectives involved in shaping them. No single consultancy can hire its way out of an industry-wide pipeline problem. But every consultancy can decide whether they treat it as someone else's to solve, or as something that shows up in every client delivery.
