Harsh Vardhan Mishra is presently serving as a Practice Head - AI. He is a seasoned data scientist and consultant with over 18 years of expertise. He is passionate about transforming complex data into actionable insights, with deep experience addressing challenges across the transportation, healthcare, and education sectors. Through extensive work in training, development, and consultation, Harsh is committed to empowering professionals and organizations to unlock the full potential of data science.Beyond his core data science practice, Harsh leads a multidisciplinary team of experts spanning social media and digital marketing, Python, machine learning and generative AI, Java development, CRM solutions, and web application development. This breadth allows him to offer innovative, tailored solutions — whether the goal is upskilling a team, building cutting-edge technical solutions, or optimizing an organization's digital presence.

Who's Accountable When AI Causes Harm? — legal and ethical accountability in an AI-mediated world                                                                                                                                                         AbstractAs Artificial Intelligence (AI) systems transition from administrative tools to autonomous decision-makers across critical sectors—including healthcare, finance, employment, and autonomous transportation—the boundary of legal and moral liability has grown increasingly opaque. Traditional legal frameworks, historically designed to assign liability to distinct human actors or corporate entities, struggle to accommodate autonomous, probabilistic, and self-learning algorithmic systems. This presentation explores the complex intersection of jurisprudence, corporate governance, and ethics to answer a central question: When an AI system causes harm, where does the legal and ethical responsibility lie? The presentation examines the current liability paradigms, evaluating the viability of existing legal mechanisms—such as strict product liability, vicarious liability, and professional negligence—when applied to complex deep learning architectures. It addresses the systemic challenge posed by the "black box" problem, where the opacity of proprietary algorithms complicates proof of causation and fault. Furthermore, the discussion analyzes emerging regulatory approaches worldwide, including risk-based compliance frameworks and proposals for mandatory algorithmic audits, strict liability for high-risk applications, and compulsory liability insurance. Beyond strict legal mechanics, the session addresses the broader ethical implications of delegating high-stakes decisions to autonomous systems. It provides practical governance frameworks for enterprise leaders, AI developers, and policy makers to establish end-to-end accountability throughout the AI development lifecycle—from training data curation and model design to deployment and continuous monitoring. Ultimately, the presentation outlines actionable strategies for fostering responsible innovation while ensuring robust legal remedies and ethical recourse when AI systems fail.