Power of AI Done Well: What needs to change?
In our last post, we explored the ripple effect of responsible AI and how values-led design can create positive change that spreads across systems, sectors, and borders. These ripple effects are a testament to what’s possible when innovation is rooted in equity, community, and care.
But to make responsible AI the global norm rather than the exception, we need more than good examples. We need change across cultures, practices, industries, and power structures.
In this ninth installation of The Power of AI Done Well, we surface the shifts our Top 33 nominees say are most urgent, and most possible to build a future where AI works for everyone.
A Mindset Shift: From Profit-First to People-and-Planet-First
The most urgent transformation? Rebalancing our priorities. Too often, AI development is driven by speed, profit, and market competition. Our nominees argue for a broader, more holistic lens, one that centres human and environmental well-being alongside innovation.
Changing this mindset isn’t just about ethics. It’s about long-term resilience and impact. When AI is guided by values like equity, transparency, and sustainability, it’s far more likely to deliver public trust, lasting relevance, and global benefit.
Rethinking Relationships: Developers and Civil Society
Another critical shift is reimagining how AI creators interact with society. Today, development often happens behind closed doors with little public input until a product is already live.
Our nominees call for a new model: one rooted in transparency, accountability, and ongoing dialogue with civil society. Open-source approaches, public interest audits, and inclusive design methods are all part of this rebalancing. The goal? Move from "build fast and fix later" to "co-create and build better."
Inclusion in Every Layer: From Code to Culture
Diversity in AI isn’t just about team photos, it’s about outcomes. Our nominees advocate for greater inclusion in:
Programming teams, to reflect varied lived experiences
Training data, to challenge historical bias
Language and culture, to end digital exclusion
Decision-making power, to shift what is prioritised and funded
More representative systems are also more innovative, more ethical, and more widely trusted.
Tackling the “Black Box” Problem
Opaque AI models, those that make decisions without clear logic or accountability, pose major risks, especially in health, education, justice, and employment.
Transparency and explainability must become default design principles. This not only builds public trust but empowers users to understand, question, and challenge AI decisions that affect their lives.
Opening the Code: From Proprietary to Participatory
Much of AI today is locked behind corporate patents and closed models. But what if AI were more open-source than anything we’ve built before?
Our nominees believe a more transparent, collaborative approach to AI development could enable shared progress, build capacity in underrepresented regions, and accelerate socially beneficial innovations.
Asking the Right Questions
Perhaps the most radical suggestion from our Top 33? Just because we can build AI for something doesn’t mean we should.
Responsible AI means asking not only how a tool works, but whether it’s needed, who it serves, and what it might displace. Taking pause to reflect first enables us to avoid gimmicks, hype, and extractive applications, and to make space for purpose-driven, community-informed solutions.
AI Literacy for All
Finally, change begins with understanding. Too few people across society, government, and even within tech fully grasp AI’s implications. Making AI literacy the norm, not the exception, is essential for meaningful participation, accountability, and democratic oversight.
From school classrooms to government briefings, we need to cultivate a shared awareness of AI as a socio-technical system shaped by power, values, and human choices.
What’s Next
What needs to change is clear and many in our global community are already making that change happen. But transformation isn’t led by systems alone. It’s led by people. And the women in our She Shapes AI network are doing just that.
In our next post, we shine a spotlight on Women as Disruptors, those challenging the status quo, bridging sectors, and redesigning the very foundations of tech leadership.
Join us for Part 10 as we celebrate bold action, fresh thinking, and a new definition of disruption, one grounded in equity, community, and courage.
