Truth or Tech: The Socio-Technical Risks of AI Audio for the Global South

Generative AI continues to blur the lines between synthetic and authentic content. While many have grown skeptical of believing everything they see, the emerging challenge lies in distinguishing between real and AI-generated audio. From robocalls and voice phishing to fabricated ransom demands, the misuse of AI-manipulated speech is on the rise. While there are existing tools to combat and address this challenge in the Global North, these solutions are often trained on narrow linguistic datasets that fail to reflect the language diversity of the global majority. Given the widespread deployment of AI-manipulated audio and deepfake technology across parts of the Global South, it is crucial to define protections for sustained resilience for those in the region. Join the Strategic Foresight Hub for a timely conversation with Responsible AI Fellow Pamposh Raina as we examine critical gaps in audio detection systems, explore how they can be strengthened, and consider what more equitable AI safeguards should look like going forward.