Harnessing AI to Empower Underserved Agricultural Communities
The rise of AI provides a unique opportunity to address some of the inequalities that exist in financial access, affecting traditionally underserved populations such as those in emerging countries (Economist 2024). The agricultural industry is one example: plagued with underserved populations, it can benefit greatly from AI. Besides the use of AI in the field, by increasing efficiency in the use of resources, for example, it has the ability to give farmers access to value chains that were formerly inaccessible to them due to market asymmetries (Cook and O’Neill 2020). By providing alternative forms of vetting through data collected from various nontraditional sources, AI has increased accessibility to financial services (Liu et al. 2023). AI automates many of the costly tasks for financial service providers (FSP), incentivizing them to allocate more resources that make credit products more accessible and less costly (Biallas and O’Neill 2020). Organizations surveyed globally indicated that 65 percent were using Generative AI (Gen AI) for at least one of their business functions on a regular basis (Singla et al. 2024).
The estimated financing gap is $1.5 trillion, which affects women and small and medium enterprises (SMEs) disproportionally (Biallas and O’Neill 2020; Kim 2019). The Asian Development Bank’s 2022 survey found that banks reportedly rejected 45% of trade finance applications of SMEs (Beck 2023). With nontraditional data, such as production records, sales, timely payments, and chargebacks, AI has the capacity to be cost-effective and “create tailored financing solutions, assess credit risk, and help predict fraud and detect supply chain threats in real time” (Biallas and O’Neill 2020, 4). However, actualizing this potential depends on how well the foundational structures are established by governments, businesses, and investors for the responsible and sustainable adoption of AI (ibid.). AI’s real-time, analytical abilities are effective with nontraditional data, bridging financing gaps with alternative credit scoring, by identifying threats to FSPs, and through the automation of business models (ibid.).
AI can provide alternative credit scoring by using nontraditional data (Liu et al. 2023). Data, such as satellite images and social media records increases the access of commonly marginalized groups, such as women, to FSPs (Biallas and O’Neill 2020). Social Lender is one example of a company using data from an individual’s phone and social media to generate credit scores (IFC-BMZ 2024). Based on that person’s “social reputation,” this score can then be used to apply for financial services (Social Lender 2024). Another example, AgroFides, leverages “world-class agricultural expertise and cutting-edge data analytics,” and agronomic data to generate credit scores for farmers in Ghana (AgroFides 2024).
The Identification of threats, such as cyber-attacks is crucial for financial institutions, making identity verifying and risk management tools invaluable (Wolf 2024). Fraud detection and risk management are the key reasons for financial institutions to employ AI (ibid). Regulatory technology companies (RegTech) use technology to help businesses comply with regulations, including those governing know-your-customer (KYC) and anti-money-laundering (AML) (Ravi et al. 2023). GenAI can impact RegTech by automating the collection and processing of data and the generation of reports, improving precision and efficiency (Prakash 2024). These advancements improve the capacity of FSPs to assess predominantly three issues related to identity verification: (1) the customer’s compliance and reputational risks, (2) proper inspection of other parties involved in transactions, and (3) the assessment of customer transactions in detecting abnormalities (Shin 2024).
Automated processes are enhanced by AI, decreasing costs for financial institutions (MIT 2024). Accessibility is increased for those who would normally not have access to such services due to language barriers or lack of physical branches (African Business Magazine 2024). Arturito, an example of such technology, provides around-the-clock customer service through Facebook and currency conversion assistance for Bank BCP in Peru (Biallas and O’Neill 2020). WorldCover is a technology that uses “satellite, weather station, and agronomic data” to evaluate and predict weather events for automatic insurance payouts (Biallas and O’Neill 2020, 5). Furthermore, those insured who do not have bank accounts may use nonbank payment providers, such as M-Pesa, for their automatic payouts (ibid.). AI has also created opportunities for collateral management services to be automated (Clearstream 2024). Clearstream’s application, Own Selection Criteria with Automated Reasoning (OSCAR), gives customers the ability to create their own custom collateral baskets (ibid.).
Personalized banking is another costly aspect of finance that can be made more inclusive with the use of AI (Biallas and O’Neill 2020). Personalized recommendations relating to loans, repayment plans, and interest rates are made by AI using financial profiles and farm data (Buchanan 2023). For example, in India, personalized AI-generated advice is given for mutual funds compatible with the customer’s data using ArthaYantra’s AI instrument: Arthos (Biallas and O’Neill 2020).
Risk and Government Response
The technology must appropriately address the challenges of those it is meant for, such as SMEs, by understanding their needs and limitations (Cook and O’Neill 2020). Therefore, investments into ensuring the consumers are well-trained and versed in the technology and resources available to them is crucial (ibid.). The responsible employment of the technology to avoid a rise in indebtedness will rely on the proficiency and training of the staff tasked with utilizing them (Biallas and O’Neill 2020). Developing “robust privacy, data management, cyber security, and supervisory regulations/processes to facilitate AI adoption” are all ways to mitigate the abovementioned risks (Biallas and O’Neill 2020, 6).
Data inaccuracies, third-party dependencies, an increased risk of cyber-attacks, and algorithmic bias and discrimination are some of the risks attributed to AI (Shin 2024). For example, women were affected negatively by AI algorithmic assessments of their creditworthiness when their gender was included in the information they had to provide (Sarah 2024). Privacy concerns have also been raised in relation to surveillance, ownership, and usage of the data (Sparrow et al. 2022). Regulations are essential in addressing a lack of competition for the consumer: another potential risk when a segment of an industry is faster to adopt this novel technology (Biallas and O’Neill 2020).
The European Data Act aims to increase accessibility to data in addition to encouraging fair competition in the data market (De Baerdemaeker 2023). Hence, data shall be regulated in the manner in which it is shared, stored, accessed, and used (European Data Act 2023). Furthermore, this law is meant to lead to top-level AI reference testing and experimentation facilities (TEF), Digital Innovation Hubs (DIH), and the enhancement of digital skills (De Baerdemaeker 2023). De Baerdemaeker (2023) does warn, however, that it is important for regulations to differentiate between high risk and low risk AI, i.e., AI dealing with personal data has been treated in the same manner as AI that collects data on crops in the fields (De Baerdemaeker 2023). The EU is not the only government taking steps to ensure a smoother transition into a world with AI. The government of Indonesia has begun the process for the standardization of data collection and sharing while addressing privacy and ethical concerns attributed to AI (Sarah 2024). So has the U.S. government, with the National AI Initiative Act of 2020 meant to oversee and implement the nation’s AI strategy (H.R. 6216 2020).
Looking Ahead
AI, GenAI in particular, will save FSPs hundreds of billions of U.S. dollars annually (MIT 2024). But the industry needs to update its technology and data storage to ensure compatibility with the ways in which AI can enhance its business functions and services (ibid.). Additionally, FSPs need regulations to guide them (MIT 2024). Coordinating with governments will foster guidelines that will protect consumers without hindering innovation (Kinyua 2024). AI will change the ability of FSPs to meet customer demands for the better, nonetheless their success hinges on their ability to successfully overcome the abovementioned challenges while complying with relevant regulations (ibid.).
References
African Banker Magazine. “Africa’s Banking Groups Embrace Disruption and Innovation in Digital Services.” African Business, March 1, 2024. https://african.business/2024/02/partner-content/africas-banking-groups-embrace-disruption-and-innovation-in-digital-services.
AgroFides. “Agrofides, Inc.” AgroFides. Accessed September 3, 2024. https://agrofides.com/.
Beck, Steven. “2023 Trade Finance Gaps, Growth, and Jobs Survey.” Asian Development Bank, September 5, 2023. https://www.adb.org/publications/2023-trade-finance-gaps-growth-jobs-survey.
Biallas, Margarete and O’Neill, Felicity. Artificial Intelligence Innovation in Financial Services, 2020. https://documents1.worldbank.org/curated/en/839801596184068790/pdf/Artificial-Intelligence-Innovation-in-Financial-Services.pdf.
Buchanan, Shelby. “AI and Agriculture: How It’s Changing Farming and Finance.” AgAmerica, January 29, 2024. https://agamerica.com/blog/ai-and-agriculture/.
Cook, Peter and O’Neill, Felicity. Artificial-intelligence-in-agribusiness-is-growing-in- …, 2020. https://documents1.worldbank.org/curated/en/198451596436781534/pdf/Artificial-Intelligence-in-Agribusiness-is-Growing-in-Emerging-Markets.pdf.
Clearstream. “Automating collateral management with Oscar.” Accessed September 3, 2024. https://www.clearstream.com/clearstream-en/securities-services/collateral-lending-liquidity/collateral-management/triparty-collateral-services-cmax-/oscar/automating-collateral-management-with-oscar–3126750.
De Baerdemaeker, Josse. Artificial Intelligence in the Agri-Food Sector: Applications, risks and impacts, January 2023. https://www.europarl.europa.eu/RegData/etudes/STUD/2023/734711/EPRS_STU(2023)734711_EN.pdf.
Economist. “AI Holds Tantalising Promise for the Emerging World.” The Economist, 2024. https://www.economist.com/leaders/2024/01/25/the-tantalising-promise-of-ai-for-the-emerging-world.
European Data Act. “The European Data Act.” The European Data Act, December 13, 2023. https://www.eu-data-act.com/.
IFC-BMZ. Scaling-up-farmer-financing-through-AGTECHS-in-sub- …, 2024. https://www.ifc.org/content/dam/ifc/doc/2024/scaling-up-farmer-financing-through-agtechs-in-sub-saharan-africa-ifc-2023.pdf.
Kim, Kijin. “2019 Trade Finance Gaps, Growth, and Jobs Survey.” Asian Development Bank, September 3, 2019. https://www.adb.org/publications/2019-trade-finance-gaps-jobs-survey.
Kinyua, Chris. “The Rise Of Artificial Intelligence In Finance.” Forbes, August 12, 2024. https://www.forbes.com/councils/forbestechcouncil/2024/08/12/the-rise-of-artificial-intelligence-in-finance/#:~:text=AI%20will%20become%20increasingly%20integrated,is%20crucial%20for%20successful%20implementation.
Liu, Yan; Boy, Hans Christian; Khurana, Saloni; and Sinha, Anshuman. Artificial Intelligence: Revolutionary Potential and Huge Uncertainties, 2023. https://openknowledge.worldbank.org/server/api/core/bitstreams/c2696968-4f2d-5477-8825-3d9073a6b8e8/content.
MIT. “Finding Value in Generative AI for Financial Services.” MIT Technology Review Insights, 2023. https://wp.technologyreview.com/wp-content/uploads/2023/11/MIT-UBS-generative-AI-report_FNL.pdf.
Prakash, Sanjeev. “The Integration of Generative AI in RegTech: Rethinking Regulatory Compliance.” Innovation & Tech Today, May 13, 2024. https://innotechtoday.com/the-integration-of-generative-ai-in-regtech-rethinking-regulatory-compliance/.
Raji, Ravi, Pradeep Ravi, and Onkar Sumant. “Regtech Market Size, Share and Growth: Forecast Report, 2032.” Allied Market Research, August 2023. https://www.alliedmarketresearch.com/regtech-market.
Sarah, Maria. “Using Alternative Data and Artificial Intelligence to Expand Financial Inclusion: Evidence-Based Insights.” The Abdul Latif Jameel Poverty Action Lab (J-PAL), March 21, 2024. https://www.povertyactionlab.org/blog/3-21-24/using-alternative-data-and-artificial-intelligence-expand-financial-inclusion-evidence.
Shin, Hyun Song. “III. Artificial Intelligence and the Economy: Implications for Central Banks.” The Bank for International Settlements, June 25, 2024. https://www.bis.org/publ/arpdf/ar2024e3.htm.
Singla, Alex; Alexander Sukharevsky; Yee, Lareina; and Chui, Michael. “The State of Ai in Early 2024: Gen Ai Adoption Spikes and Starts to Generate Value.” McKinsey & Company, May 30, 2024. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai#/.
Social Lender. Accessed September 3, 2024. https://www.sociallenderng.com/.
Sparrow, Robert, Mark Howard, and Chris Degeling. “Managing the Risks of Artificial Intelligence in Agriculture.” NJAS: Impact in Agricultural and Life Sciences 93, no. 1 (January 1, 2021): 172–96. https://doi.org/10.1080/27685241.2021.2008777.
H.R.6216 – 116th congress (2019-2020): National Artificial Intelligence Initiative Act of 2020 | congress.gov | library of Congress. Accessed September 3, 2024. https://www.congress.gov/bill/116th-congress/house-bill/6216/text.
Wolf, Brett. “Ai, Other Technology the ‘Only Answer’ to AML Challenges in Evolving Threat Landscape, Says Acams Report.” Thomson oneills Institute, February 20, 2024. https://www.thomsonreuters.com/en-us/posts/investigation-fraud-and-risk/ai-aml-challenges-acams/.
