Case Study 1: AI Use Case(s): DPG & DPI
Introduction
Digital inclusion, in the 21st century, is a concept that operates under the guise of Industry 4.0. The Henry Ford notion of customers adjusting their tastes to the product is no longer feasible; products must be flexible and fit following the needs of the customer to provide a breeding ground for subsequent innovation with the holistic view of social progress. To allow innovation, data is considered an important part and lifeblood of Industry 4.0.
The Industrial Revolution 4.0 has shaped our lives and transformed how we connect, it wouldn’t be wrong if we would characterize this phase with the rampant use of digital servitization. It can also be remembered with certain other buzzwords such as DPI (Digital Payment Infrastructure), DPG (Digital Public Goods), UPI (Unified Payment Interface), cloud computing, artificial intelligence, and blockchain, among others.
What is Digital Public Goods (DPG) & Digital Public Infrastructure (DPI)?
DPG is an Open Source Software, Data, AI Models, standards and content that adheres to privacy regulations and other applicable laws and best practices while aiming to achieve the Sustainable Development Goals (SDGs) as recognized by the United Nations (UN) and certified by the Digital Public Goods Alliance (DPGA). DPI is the realization of the value of DPGs in population-scale implementation. Foundational DPI facilitates three kinds of flow(s) i.e. People Flow: Through Digital ID Systems, Money Flow: Via Real-Time Payment Systems, and Information Flow: Through consent-based data-sharing systems. An effective DPI foundation in place can empower citizens to control their data while promoting equitable access to public infrastructure, with security and future ecosystem development front of mind.
India Stack: A trailblazer in foundational DPI
India, through its Indian Stack, has become the first country to develop all three foundational DPIs: digital identity (Aadhar), real-time fast payment (UPI), and a platform to safely share personal data without compromising privacy (Account Aggregator built on the Data Empowerment Protection Architecture or DEPA). Each DPI layer fills a clear need and generates considerable value across sectors. When aggregated, foundational DPIs constitute the backbone of a country’s digital infrastructure. These layers interface with each other to create an ecosystem that facilitates seamless public service delivery and allows businesses to design novel solutions on top of the DPI layers. In turn, this enables the creation of Open Networks as not seen before. India is now developing such open networks for other emerging challenges including but not limited to credit (Open Credit Enablement Network), commerce (Open Network for Digital Commerce – already rolled out), Open Health Services Network (UHI), and many more.
The G20 New Delhi Leader’s Declaration also reaffirmed the role that the Digital Public Infrastructure (DPI) would play in advancing financial inclusion under the cornerstone of inclusive growth and sustainable development. The impact of DPI goes beyond inclusive finance—it can support health, education, and sustainability. Amid the COVID-19 pandemic, DPI enabled emergency support to be directly delivered to the digital wallets of those in need and helped facilitate swift vaccine distribution. The India Stack exemplifies this approach, combining digital ID, interoperable payments, a digital credentials ledger, and account aggregation. As mentioned in the report “G20 Policy Recommendations for Advancing Financial Inclusion and Productivity Gains through DPI”, the World Bank stated that India has in just six years achieved a remarkable 80% financial inclusion rate—a feat that would have taken nearly five decades without a DPI approach.
Scaling India’s Success across the APAC region
To this effect, India has created the Modular Open Source Identity Platform (MOSIP), presently adopted by 11 nations with more than 100 million active users. We also need to develop global standards through a multilateral dialogue led by India. If standards originating from developed nations were transplanted to an emerging economies’ context without deferring to their developmental concerns, smaller countries would simply be captive to dominant technology players. For India’s DPI success to become a worldwide revolution, three types of institutions must be built. First, we need independent DPI steward institutions to have a governance structure that is agile and responsive. A multiparty governance process through independent DPI institutions will be accountable to a broad range of stakeholders rather than be controlled by a single entity or group. This can build trust and confidence in DPI. Additionally, Big Tech would likely engage in regulatory arbitrage to concentrate power without these standards. Finally, we need to develop sustainable financing models for developing DPI for the world.
Case Study 2: Artificial Intelligence in Targeted Advertising
Introduction
The proliferation of advanced technology, such as artificial intelligence (AI), has transformed how organizations process personal data. Organizations leverage this emerging technology to collect and analyze large amounts of data to enable automated decision-making and profiling, leading to more accurate and precise business decisions. The data becomes highly valuable, and organizations can capitalize on these opportunities. This emerging technology is gaining rapid acceptance and penetrating various industries, including healthcare, education, financial services, and marketing (ICO, n.d.).
Marketing techniques and consumption patterns change as customers’ preferences and expectations evolve. Instead of waiting for the customers to approach the marketplace, they can reach potentially profitable customers anytime and anywhere (Selbst, 2017).
Targeted advertising is a business strategy that sends a tailored advertisement to persuade a person or a specific group to buy specific products. Instead of waiting for customers to come to them, businesses can reach out to customers anytime and anywhere with personalized recommendations and curated access to retailers (Jokubaitis, 2018).
Facebook is one of the biggest online advertising platforms with an extensive reach of users and many advertisers. Facebook has introduced new and innovative methods to target users more effectively by collecting a significant amount of personal data about their users rapidly and continuously from disparate devices and sources, then made available to advertisers. This technique has been adopted by other social media and social networking platforms (Till Speicher et al., 2018).
Risks and Challenges
Targeted advertisement has several risks and challenges stemming from its opaque AI systems and complex decision-making processes that are not easily explainable. It is hard for customers to challenge algorithmic decisions when they don’t understand how it works(Hassija et al., 2023).
Targeted advertising has powerful incentives to look for ways to help and delight its users, but at the same time, it has incentives to look for ways to exploit users. The advertisements could influence purchasing decisions (Rebecca Walker Reczek et al., 2016) and political campaigns, like how Trump won in the 2016 US election (BBC News, 2020). This technique potentially reinforces societal inequality by excluding sensitive groups from housing, employment, and financial advertisements(Till Speicher et al., 2018). Even more disturbing are the advertisements that contain hate speech, incitement, and promote violence (Biddle, 2023).
From a legal perspective, it is considered a violation of privacy. Users feel they lack control over their data for intrusive data collection and processing for unnecessary purposes. They would find it objectionable if they were aware of the practice (Calo, 2013). Moreover, even though there are privacy policies, they are often too complicated to understand, containing many legal jargon, and users may unknowingly consent to a legally binding document.
The intersection of DPI and AI in Targeted Advertising
The development of Digital Public Infrastructure (DPI) in India, now encompassing people flow, money flow, and information flow, has significantly transformed the landscape of data collection, storage, and utilization. These innovations have opened new avenues for AI algorithms in targeted advertising by providing access to a wealth of consumer data, ranging from biometrics to financial transactions. This access can facilitate highly personalized marketing strategies and improved consumer outcomes, however, utilizing DPI for targeted advertising and AI integration introduces substantial challenges concerning privacy, security, and the ethical use of data.
Opportunities and Challenges
DPI systems furnish AI with a rich dataset, enabling the derivation of profound marketing insights. This capability supports the creation of personalized advertising strategies that can drastically improve consumer engagement and business growth. By integrating DPI with AI algorithms, advertising can be more closely aligned with the consumer’s preferences and needs, thereby increasing efficiency and satisfaction.
The extensive and sensitive nature of data managed within DPI systems demands advanced security measures to prevent data breaches and misuse, thus safeguarding individual privacy.
Utilizing DPI data for targeted advertising raises complex ethical questions regarding consent, transparency, and the extent of data use. Addressing these issues necessitates comprehensive policy frameworks (Agbese, Mohanani, Khan, & Abrahamsson, 2023; Samarawickrama, 2022).
Frameworks and Solutions
The adoption of international standards for data such as the EU GDPR and AI ethics is crucial for establishing a secure and ethical framework. These standards should encompass ethical, technical, and policy aspects to protect consumer rights, privacy, and transparency. Samarawickrama (2022) underscores the importance of AI governance and ethics frameworks for sustainable AI, advocating for social diversity, equity, and inclusion as essential for mitigating risks throughout AI development and fostering social justice.
Blockchain technology offers a promising avenue for bolstering security and data governance within DPI systems. Its characteristics of decentralization, immutability, and transparency can significantly improve data integrity and management (Pournaras, 2020). Pournaras (2020) introduces the concept of “Proof of Witness Presence,” leveraging blockchain consensus to enhance citizens’ trust and address security vulnerabilities, which could be instrumental in strengthening the security, privacy, and transparency of DPI systems.
Aligning technological advancements with the Sustainable Development Goals (SDGs) is essential for fostering a technology landscape that is not only advanced but also equitable and sustainable. This alignment highlights the broader objectives of social and economic development beyond mere technological innovation.Conclusion
The integration of Digital Public Infrastructure with targeted advertising and AI algorithms presents a complex landscape filled with both opportunities and challenges. While there is significant potential for personalized marketing and enhanced consumer experiences, the ethical, security, and policy implications cannot be ignored. Embracing international standards, exploring innovative solutions like blockchain, promoting responsible AI use, and aligning with broader development goals are critical steps toward navigating this intricate terrain. Multilateral, collaborative efforts are essential in developing a unified approach that ensures technological advancements benefit society at large, privacy is safeguarded and developments are made in a sustainable manner.
Written by Sameer Gahlot, Nur Adlin Hanisah Binti Shahul Ikram, Jack Kelliher, Caresse Tan Zhi Nie (Edited by Bea Guevarra)