The rapid integration of artificial intelligence into financial markets is transforming the architecture of investment decision-making. As algorithmic trading systemsoperate at increasing speed, scale, and autonomy, systemic risk may emerge notfrom irrational human behavior but from synchronized algorithmic rationality. Thispaper asks a critical question: Can AI trigger the next financial crisis, and if so,who bears responsibility?The study introduces the concept of a structural responsibility gap in AI-mediatedinvestment environments. By distinguishing computational output from normativejudgment, it argues that investment decisions inherently involve risk endorsement,value commitment, and accountability beyond probabilistic calculation. Whenjudgment is delegated to autonomous systems, responsibility becomes layered andfragmented across developers, institutions, investors, and regulators.Through analysis of algorithmic feedback loops, synchronization dynamics, andcrisis amplification mechanisms—including flash crash events—the paper demonstrates how automated coordination may intensify systemic instability. To addressthis gap, it proposes a Hierarchical Joint Responsibility Model designed to governdelegated judgment without attributing moral agency to AI systems. The stabilityof algorithmic finance ultimately depends not solely on technological control, buton the structural redesign of accountability