Levelling the AI playing field for women in agriculture

Artificial intelligence (AI) is touted as a game-changer. Companies like Neurafarm in Indonesia have developed AI-powered solutions to detect plant diseases and recommend treatment solutions for farmers, while Ignitia in Ghana has developed climate intelligence and forecasting solutions for agribusinesses, providing much-needed support for smallholder farmers.

However, these advancements risk perpetuating gender biases if women farmers are not included in the data used to develop and test such technologies. Most leading AI companies are led by men and 70 percent of the sector’s workforce is male.

Additionally, AI models for agriculture often rely on training data that’s been collected primarily from male farmers.

This means the algorithms are designed to solve problems from a male perspective, leaving out the farmers such as access to land or the double burden of unpaid domestic labor unique challenges faced by women.

Without targeted interventions, the gender digital divide will continue to grow, leaving women farmers at a disadvantage in adopting new technologies like AI. So how can we ensure AI benefits women as much as it does men?

Read: Agriculture dominates AI use cases in Kenya, report shows

Design inclusive AI

First, we need gender-balanced data to train AI models to eliminate any misrepresentation. Developers must make sure that the training data includes inputs from both male and female farmers to reflect the diverse challenges they face. Involving women in the design, testing, and development of AI tools will help create solutions that are not only technologically sound but also socially inclusive and acceptable.

Open data

Second, making agricultural data freely available can encourage the development of more inclusive AI technologies. Stakeholders need to make data freely available for all people and entities to use.

Governments, tech firms, and women’s organisations should collaborate to ensure that both men and women benefit equally from these innovations. Open data initiatives can also help track gender disparities in agriculture, providing actionable insights for policymakers.

AI governance structures

Research conducted by the Stanford Social Innovation Review indicates that 44 percent of existing AI systems globally exhibit gender bias.

There is a need for well-developed and validated AI governance models that have the potential to prevent and address any racial or gender bias exhibited by AI solutions and be able to reinforce safety and privacy standards.

Targeted education and training

Access to AI is not enough if women do not have the skills to use it. Equipping women with essential skills and knowledge to adopt AI solutions and participate fully in the field is key.

Women, especially in rural areas, need targeted digital and AI literacy training alongside skills in data management, using AI technologies in agriculture, empowering them to fully participate in the digital revolution, levelling the playing field.

This is where government programs and private sector outreach to farming communities become crucial.

Training women alongside men builds value for diverse perspectives, boosting women's participation in technology and enhancing the overall efficiency and productivity of agricultural systems.

Increase women’s participation

Finally, it is essential that women are represented in leadership roles within the AI and agricultural sectors. Only about 18% of leadership positions in AI are held by women, a glaring gap that needs addressing.

By increasing women’s participation at the decision-making level and building diverse teams, we can ensure that AI solutions are more gender-responsive and better equipped to tackle the complex challenges in agriculture.

Further, women’s inherent skills in communication, empathy, etc. may be used in generative AI for agribusinesses to include women in market or customer outreach functions and design more gender-responsive solutions.