By Elizabeth (Liz) Ngonzi

As decision-makers, enterprise leaders, and policy architects gathered during United Nations General Assembly (UNGA81) week in New York, the conversation around Artificial Intelligence reached a critical tipping point. We are moving rapidly past initial AI hype into the sober reality of institutional adoption. Yet, a fundamental contradiction lies at the heart of today’s AI economy: while generative technology is advancing at exponential speed, the way many organizations deploy it remains dangerously extractive.

When organizations view Generative AI strictly through a lens of cost-cutting, headcount reduction, or automated shortcuts, they do not just risk operational fragility and loss of institutional memory—they scale historical biases and structural inequities at algorithmic velocity.

Recent global workforce and healthcare data unveiled during the World Woman UNGA Agenda convening in New York City laid bare this divide across two pivotal sessions I was privileged to attend:

Women’s Health & Equality Moonshot: Highlighting how historical medical research and foundation LLM datasets overwhelmingly default to male biological parameters, risking "garbage in, garbage out" diagnostic errors unless data is systematically sex-disaggregated.

Women at Work & Equality Moonshot: Revealing that while global demand for AI skills has doubled over two years, women account for only 25% of new AI hires globally, hold just 13% of technical/leadership roles in AI firms, and face a 28% global leadership progress wall that has remained flat since 2022.

As I recently noted in CNET:

To bridge these gaps, we must stop treating technology as a tool for human replacement and start designing Generative Intelligence as an engine for human capability, equity, and institutional agency.

(Reflecting on key insights from the first session featuring Dr. Andrea Feigl, CEO of Health Finance Institute; Wendy Lund, Global CEO of Allison Worldwide & Vice Chair of Health at Stagwell; and Seema Kumar, CEO of Cure)

While my schedule only allowed me to take in part of the Women’s Health & Equality Moonshot panel, the proceedings underscored an inescapable truth: in healthcare and life sciences, generative AI systems are only as reliable as their underlying data layer.

For decades, clinical trials and medical studies treated male biology as the universal default. If open-source LLMs, diagnostic AI tools, and foundation health models inherit these skewed historical datasets, they will automate medical misdiagnosis at scale.

To ensure AI serves as a healthcare equalizer rather than an engine of compounded inequality, institutional leaders and policymakers must enforce three non-negotiable governance pillars:

Four Concrete 2030 Moonshot Benchmarks:

(Synthesizing findings from the panel featuring Sarah Steinberg, Head of Global Public Policy Partnerships at LinkedIn; Kiersten Barnet, Executive Director of the New York Jobs CEO Council; and Françoise Cardoso, Global Director of Corporate Social Responsibility at L-Acoustics)

The workforce session brought the structural realities of the AI labor market into sharp focus. Drawing from LinkedIn’s global economic graph, Sarah Steinberg shared sobering metrics:

Addressing these disparities requires a fundamental shift in corporate strategy. As I stated in my interview with TechRound:

Operationalizing

1+1+AI=10™ Throughout my 30-year trajectory across technology transformation, management consulting, and executive education at NYU, I reframe this challenge through my signature 1+1+AI=10™ methodology:

Extractive AI asks: "How many headcount can we eliminate?"

Regenerative Intelligence asks: "How much new human potential, strategic judgment, and resilience can we unlock?"

Case Study in Regenerative Reskilling: Northwell Health A standout case study shared by Kiersten Barnet (representing New York's 25 largest private employers and 250,000 CUNY students) is Northwell Health. Facing the inevitable automation of telephone insurance claims processing, Northwell did not displace those workers—roles overwhelmingly held by women. Instead, they proactively retrained them to become certified, high-value mammographers, directly solving a critical healthcare talent shortage while preserving employee earning power.

As Françoise Cardoso highlighted from L-Acoustics' corporate strategy, caregiving responsibilities (childcare, eldercare, mental load) must be integrated as core operational performance factors rather than treated as soft corporate perks. Equal, mandatory parental leave and co-designing AI tools with female employees ensure that technology enhances human capability rather than inducing burnout.

While demand for technical AI prompt engineering has grown by 70% YoY, enterprise recruiters and C-suites increasingly recognize that technical tools commoditize rapidly. As I shared during my feature in Building Creative Machines:

The fastest-growing, highest-value skills in an AI-driven economy are distinctly human:

Modern leadership demands non-linear, cross-functional, and interdisciplinary backgrounds—a profile that women and mission-driven leaders overwhelmingly possess. In an economy flooded with cheap synthetic output, human judgment remains un-automatable.

Culture is not a mission statement on a corporate wall; culture lives in your daily operational mechanics. Does your technology free human energy, or does it extract it?