Modern distributed systems increasingly depend on workflow orchestration frameworks to coordinate long-running operations across loosely coupled services. Platforms such as Temporal and Cadence have significantly improved reliability, fault tolerance, and operational consistency by introducing workflow-as-code execution models capable of handling retries, state persistence, and deterministic replay. These systems transformed orchestration from fragile procedural coordination into durable distributed execution. However, as distributed environments continue evolving, workflow orchestration faces a growing structural limitation. Existing orchestration engines generally assume that execution paths can be defined in advance through deterministic workflow logic. In practice, distributed systems operate under continuously changing conditions involving evolving services, unstable dependencies, changing APIs, shifting business rules, and unpredictable operational environments. Static workflow definitions therefore become increasingly difficult to maintain as system complexity grows. This paper examines the limitations of deterministic workflow orchestration models and introduces the concept of Intent-Driven Orchestration (IDO) as a new abstraction for distributed workflow management. Instead of defining workflows primarily as fixed sequences of execution steps, Intent-Driven Orchestration models workflows around desired operational outcomes. Under this framework, orchestration systems continuously evaluate current system state and dynamically determine execution strategies capable of progressing toward intended outcomes under changing conditions. The study explores how intent-oriented orchestration improves adaptability, resilience, maintainability, and operational flexibility within large-scale distributed systems. It further analyzes the implications of adaptive execution planning, emergent workflow behavior, observability evolution, and trust management in autonomous orchestration environments. By separating what systems must achieve from how execution occurs, this work proposes a new architectural direction for workflow orchestration systems operating within increasingly dynamic and unpredictable distributed infrastructures.