Women in Climate and AI: Bridging the Innovation Gap

Contents

Executive Summary

This panel discussion examines the critical intersection of women's economic empowerment, climate action, and AI innovation, with emphasis on systemic barriers and solutions for women-led startups in South and Southeast Asia. Panelists from Salesforce, IFC, UN Women, and climate tech startups present evidence-based strategies for integrating gender intentionality into climate and AI solutions while addressing persistent funding gaps ($1.7 trillion credit gap for women) and data biases in AI systems.

Key Takeaways

Gender Intentionality Is Not Optional—It's Infrastructure: Tools, data practices, and investment criteria must embed gender from day one. This requires funding, deliberate design choices, and measurement systems (KPIs, guardrails) built into products and processes.
Data Gaps Are Both a Problem and Opportunity: The absence of gender-disaggregated, inclusive datasets is a barrier, but also a market opportunity. Fintech companies, for example, are discovering that behavioral and cash-flow data can prove women entrepreneurs are creditworthy without collateral—solving both access and profitability challenges.
AI Solutions Must Have Women in the Loop: From data collection (who gathers data) to model design (who trains the system) to deployment (who uses it), women must be represented. Resilience AI's example of combining community volunteers with machine learning demonstrates that "human-assisted loops" prevent blind spots.
Scaling Requires Ecosystem Thinking, Not Just Individual Innovation: One startup, one tool, or one bank policy alone is insufficient. Success requires alignment across government policy, fintech infrastructure, skills training, local talent pipelines, research ecosystems, and accountability mechanisms—and this must happen regionally to enable rapid adaptation.
Men's Active Participation Is the Missing Piece: While systemic barriers for women were extensively discussed, one panelist noted that how men actively contribute to gender equity is the "largest missing piece" rarely addressed in these conversations—a critical blind spot for ecosystem approaches.

Key Topics Covered

Gender Representation & Workplace Equity: Pay gaps, glass ceiling effects, and representation across organizational levels
Climate Finance & Credit Access: Women-led entrepreneurship funding gaps; de-risking mechanisms for gender-focused investments
Gender-Intentional AI Development: Data collection, bias reduction, and inclusive AI model design
Regional Scaling of Climate-Tech Solutions: Cross-country replication and adaptation of AI solutions for climate resilience
Ecosystem Building: Multi-stakeholder coordination (government, private sector, nonprofits, fintech) to support women entrepreneurs
Data Quality & Gender-Disaggregation: Role of gender-inclusive data collection in policy-making and investment decisions
Skills Development & Access: AI literacy and capacity building for women across regions
Natural Disaster Risk Assessment: Real-world application of AI/ML for vulnerability mapping with gender lens
Philanthropic Role & Tool Access: Subsidized technology access for nonprofits; funding gaps in gender-intentional innovation
Bias in AI Systems: Preventing transfer of human gender biases into machine learning models

Key Points & Insights

Workforce Gaps Drive Product Gaps: Gender imbalance in organizational leadership directly affects product design, decision-making, and who benefits from solutions. Women comprise only 30% of the labor force in South Asia, with even lower mobile phone access rates.
Credit Access Is Severely Constrained: Of India's 64 million MSMEs, only 20-25% are women-owned, and 90% rely on informal credit. Less than 3% of venture capital funding globally goes to women-led startups; climate and tech startups fare even worse.
Data as Foundation for Accountability: NetZero Cloud and similar tools aggregate gender data (representation, pay gaps, promotion rates across levels) to make implicit biases visible and enable evidence-based decisions rather than assumptions.
Gender-Tagging in Finance Creates Behavior Change: IFC requires 60% of investments to be "gender-tagged," meaning capital deployment is tracked with specific KPIs (women employment, leadership representation, safety measures, technology access, credit access) to prevent "gender washing."
AI Has Dual Potential: AI can either deepen gender inequalities (by encoding existing biases) or reduce them (through inclusive data collection, bias detection, and cash flow-based lending that replaces collateral requirements). Outcomes depend on intentional policy design and capital flow direction.
Climate Impacts Are Gendered: UN Women research shows that in worst-case climate scenarios, 158 million more women and girls may be pushed into poverty globally compared to ~80 million men, demonstrating disproportionate vulnerability.
Inclusive Data Collection Requires Upfront Investment: Including women in data collection, ensuring gender-disaggregated datasets, and "ground-truthing" models with community input is costlier but essential. This cannot be treated as an add-on; it must be designed in from inception.
Regional Coordination Enables Faster Scaling: Solutions developed in one country (e.g., Resilience AI in India) can scale across South Asia and Southeast Asia through regional platforms (ASEAN groupings) if policies align and local talent pipelines exist.
"Know Where to Look" Is Critical for Startups: Founders must identify leverage points in ecosystems—for climate tech, this means targeting vulnerable hotspots (schools, anganwadis, community centers) and key economic sectors (hospitality, SMEs, value chains).
Institutional Solutions Must Be Embedded Early: Since climate and AI ecosystems are still nascent in many developing regions, frameworks for gender equality, ethical AI guardrails, and accountability mechanisms must be incorporated during infrastructure design, not retrofitted later.

Notable Quotes or Statements

On Hidden Biases (Puja, Salesforce): "A lot of these are implicit biases which don't even get acknowledged until there is a database supporting it."
On Gender Washing (Shalak, IFC): "We don't want gender washing here. That's why tools like yours are critical—to develop detailed KPIs around how many women-owned jobs you created, how many women in leadership you created."
On AI's Dual Nature (Puja, Salesforce): "AI also has the potential to deepen the gender gap. It is truly a matter of where capital flows at the end of the day and what kind of policies are made."
On Climate's Gendered Impact (Amar, UN Women): "About 158 million more women and girls globally may be pushed into poverty as a direct result of climate change... the number for men is almost half of this."
On Designing for Inclusion (Samita, Resilience AI): "The true definition of inclusivity is: teach the machine beforehand [about vulnerability, gender impact] rather than making it correct afterward."
On Ecosystem Reality (Samita, Resilience AI): "We all understand and appreciate ecosystem. We are an outgrowth of an ecosystem. But we don't execute it very well. The trick is in the doing."
On Scalability & Knowledge (Samita, Resilience AI): "One must know where to look... even if it is a small and medium crafts person, she or he has to know where to trade, how to source, how to sell."

Speakers & Organizations Mentioned

Speaker	Role	Organization
Zooie (moderator)	Moderator	[Unspecified]
Puja	Product/Solutions Lead	Salesforce
Shalak Tundan	Regional Head, Operations & Climate Change	IFC (International Finance Corporation)
Amar	Regional Programming Lead	UN Women (United Nations Entity for Gender Equality)
Samita	Co-founder	Resilience AI
—	—	UN Global Compact (framework partner)
—	—	ASEAN (regional grouping)
—	—	Kalari Capital (women-focused venture fund)
—	—	Vanicola (women-focused venture fund)
—	—	Colossa Ventures (women-focused venture fund)
—	—	HDFC Bank (credit profile data reference)

Technical Concepts & Resources

AI/ML Tools & Frameworks

NetZero Cloud (Salesforce): Sustainability tracking platform with gender-disaggregated workforce analytics (representation, pay gaps, promotion rates across levels)
Agent Force (Salesforce): Customizable AI tool enabling weighted investment scoring for gender-focused capital allocation; can enforce guardrails (e.g., "Have you invested in enough women-led organizations this year?")
Constitutional AI (Anthropic reference): Framework using predefined rules/principles to guide AI outputs and reduce bias
Resilience 360 (Resilience AI): Natural disaster risk assessment tool using machine learning with human-assisted loops for vulnerability mapping; operates at 96% accuracy; recognizes vulnerable hotspots (schools, anganwadis, elderly, women, children) and calculates socioeconomic asset-level losses
Sentiment Analysis & EQ Analysis: Emerging research approaches for capturing non-hard data and identifying biases in AI systems

Data & Methodologies

Gender-Tagging (IFC Framework): Systematic labeling of investments to track gender-specific outcomes across KPIs
Ground-Truthing: Community-based survey validation of AI model outputs; combines quantitative predictions with qualitative community data
WEBS Framework (UN Women & UN Global Compact): Seven-principle framework for evaluating women's participation in leadership, workforce, marketplace, and community
Cash Flow-Based Lending: Fintech approach leveraging behavioral and UPI payment data instead of collateral to assess creditworthiness
Proxy Data Usage: Using imperfect but correlated data sources (e.g., temperature proxies for urban heat island effect) when direct measurement is unavailable, validated to 95% accuracy

Research & Evidence

UN Women Climate Research: Documents 158M women and girls at poverty risk; 240M facing food insecurity; compared to ~80M men
Credit Gap Data: $1.7 trillion credit gap for women globally; <3% of VC funding to women-led startups; investment in women-led tech/climate startups at 8% (consumer business only, lower in complex sectors)
MSME Census Data: Women own 20-25% of 64 million Indian MSMEs; 90% rely on informal credit
Housing Loan Data (HDFC Bank): Women borrowers show better credit and repayment profiles than men (mentioned as evidence for fintech credit redesign)

Datasets Mentioned

Census Data: Gender-disaggregated census information
MSME Census: Small and medium enterprise demographic data by gender
Behavioral & Payment Data: UPI, mobile phone, and digital transaction histories (for credit scoring)

Policy & Strategic Implications

Regional Policy Alignment: ASEAN and similar regional groupings should develop guidance on integrating gender into climate and AI policies
Mandatory Gender Metrics: Investment frameworks should require gender-tagged capital and measurable KPIs (not "nice-to-have")
Foundational Infrastructure for Emerging Ecosystems: Bias prevention, data governance, and ethical AI guardrails must be designed into nascent climate/AI ecosystems in developing regions, not retrofitted
Skills Investment in Gig Workers & Informal Sector: Digital literacy and AI skills training must reach women in informal employment (artisans, construction, hospitality workers) who are disproportionately affected by climate disasters
Philanthropic Funding Role: Unrestricted capital is needed for expensive upfront work (inclusive data collection, community engagement, research) before commercial viability is proven

Identified Gaps & Challenges

Data Scarcity: Gender-disaggregated climate data, credit behavior data, and disaster vulnerability data remain limited in many regions
Cost of Gender Intentionality: Building gender-inclusive models, data collection, and community validation is expensive and not always viewed positively by investors
Talent Pipeline: Women in complex fields (climate tech, AI, fintech) remain scarce; retention is challenging
Bias Transfer Risk: Existing human biases in datasets risk being encoded into AI systems without deliberate countermeasures
Ecosystem Execution: While stakeholders understand the need for "ecosystem" coordination, actual implementation and accountability remain weak
Missing Male Participation: How men actively engage in gender equity solutions is largely absent from discussion
Fintech Data Governance: Privacy and consent frameworks for using behavioral/transaction data for credit scoring need clarification