By Henri Estramant
In spite of the progress made in the throughout the 20th century, many people worldwide remain unable to access banking services in their home countries, or local communities. While the trend may sound like something typical for developing countries, in the US alone 8.7 million, or 14.1% of US persons are underbanked, or unbanked according to Forbes.
The figures worldwide according to the World Bank are approximately one-third of adults , or 1.7 billion people lacking appropriate access to financial services, and basic banking. AI and Blockchain technologies could facilitate a less biased approached within the financial industry; as data shows that in many developed economies access to banks are a particular struggle for ethnic, or racial minorities, or those belonging to groups that have historically faced societal discrimination. Women that account for half of the world population, often face such challenges according to the OECD.
Advantages for the Global South
In Brazil for instance, the largest economy in South America, people of African origin remain a disadvantaged group, and one that lacks proper access to financial services. Solutions to the latter problem are being offered by São Paulo-based company Conta Black implementing a software that looks into a person’s credit score based on parameters that are relevant for those who are living in disposed communities (such as mobile phone usage data, social media data) for loans, or mortgages.
Zero-knowledge range proof solutions are likewise available through Blockchains, and are being tested by ING. As the race for regulating AI & Blockchains intensifies, countries and political entities ought to ensure facilitating the implementation of AI and Blockchain solutions for those communities and peoples that are the ‘have-nots’. End of May, the Deputy Governor of the Reserve Bank of India (RBI), Mahesh Kumar Jain, called upon the country’s banking system to adopt and AI and Blockchain technologies “to ensure sustainable growth and stability”.
In short, AI and Blockchains can be applied to anonymize names, ethnicity, gender, age, sexual orientations, etc., of those seeking access to mobile banking, loans, or mortgages; thereby blinding human prejudices. While we are not living in an ideal world, an early adoption should encourage organic growth for those technologies within the financial industry. Chatbots for instance could be trained into endemic languages of the Americas (not everyone’s native tongue is Spanish), African or Polynesian languages in order to educate them about financial services, and how best to use them. Remittances could be wired much faster and securely, and be taxed in accordance to local regulations.
Race for regulation is heating up
Any potential regulation for AI should entrench, and facilitate the application of the aforesaid technology for disadvantaged communities. As U.S. President Biden, and other developed states have recently supported membership into the G20 group for the African Union, African leaders ought to heed the call, and invite themselves as observers in already existing political structures that will shape the regulation of AI (for instance U.S.-EU Trade and Technology Council -TTC); particularly since the US and China are leading the technological development whereas the EU is leading the way in regulating, meaning that other countries and political entities, likely shall model their own legislation upon a successfully passed AI Act as being put forward by the European Union.
The G7 already has decided to pursue “trustworthy AI, in line with our shared democratic values”; developing nations are at risk of once more falling behind, if they do not ask support for AI technology implementation for their amelioration of their financial services.
EU’s AI Act as paradigm?
Algorithms and datasets should be produced responsibly in Latin America, Africa, Asia as well as the South Pacific, or otherwise they will lack the sensitivity and diversity necessary. Data and Large Language Models (LLMs) from wealthy nations cannot be applied to the Global South, albeit their learning scenarios are also relevant in developing economies.
For instance, in the European Economic Area (the EU plus Norway, Iceland and Liechtenstein) the usage of postal codes for supervised data learning is not permissible as per the “fairness” principle of the General Data Protection Regulation (GDPR – articles 39, 45, 71), as doing thus will place people hailing from those disadvantaged areas at risk of never been granted a loan as an example.
It will come as a surprise that AI is actually being adopted swiftly in many countries of the Global South, where social inequality is rampant. The more important for large Western companies to ensure they are represented in these regions, and ensure their technology is applied for legitimate purposes, and sustainable development.
About the writer:
Henri Estramant, LLM is a former consultant at the Panel for the Future of Science and Technology of the European Parliament. He is an expert in AI & Crypto regulation – certified in Conversational and Deploying AI.