The Complexities of AI for Social Good: An Explorative Study on Adversarial Machine Learning to Facilitate...

The proliferation and pervasive use of artificial intelligence (AI) pose significant challenges to our democracies. In particular, AI leverages microtargeting political campaigns by constructing detailed user profiles and inferring people's individual susceptibilities from their data. This capability enables highly targeted political messaging that can substantially influence voting decisions, potentially undermining citizens' ability to make deliberative decisions in elections. While existing research has proposed interventions to mitigate the negative impacts of political micro-targeting, many of these interventions may become less effective as AI continually yields more powerful and subliminal forms of micro-targeting. Assuming that technologies can play a role in mitigating these negative impacts of AI, we conducted an explorative study to identify the design principles of technological tools that could facilitate citizens' deliberative decision-making in elections in the continuously levelaged AI-based political campaigns. Using Indonesia's 2024 elections as a case study, we interviewed twenty citizens and four political actors to gain critical insights into how such technologies might be developed. Initially, we anticipated that privacy-enhancing AI, used to counterattack profiling AI used by political actors, might suffice to facilitate election deliberation. However, our findings reveal a more complex reality: addressing societal issues through technology is inherently challenging; no single solution can serve as a silver bullet. Instead, facilitating election deliberation requires integrating privacy with other conditions, including self-reflection, education, access to diverse information, critical thinking, and openness to others. These design principles might serve as concrete, actionable design principles to guide the development of technologies to enhance election deliberation.