As an AI researcher, I believe that both developers and deployers should share liability for AI-caused harm. Developers need to ensure their systems are safe and reliable, while deployers must be responsible for the contexts in which they apply these technologies. Collaborating on ethical guidelines and risk assessment is essential.