Generally considered to be self-learning and autonomous, AI uses IoT devices to collect and analyse data for predictive decisions. Due to its invisibility to the end users, opacity of decision-making process and algorithmic or data driven failures, exacerbated by their interconnectedness, a plethora of legal issues arise when AI fails to act in a way that is expected. This unpredictable nature, further complicates the assessment of potential harm done by AI and IoT devices, and dampens the process of assigning liability by the victims to any one entity to claim compensation for their injuries. In-spite of the global trend towards regulating AI, investments regarding development of new foundational models have grown exponentially in the recent years. This paper, doctrinally analyses if the concept of ‘Strict Liability’ can continue to be a viable regulatory framework for AI and IoT environment, or if the concept of ‘Shared Liability’, can be used to create a regulatory system, which fosters innovation and safeguards human protection at the same time, by analysing both Indian and international legal frameworks and judicial pronouncements. In the end, this paper suggests the concept of ‘Insurance’ as a hedge system to guarantee that victims receive compensation and that the companies can continue innovation in the field of AI technologies without the fear of suffering heavy losses paid as compensation for damages caused by unforeseen AI and IoT decisions.