This chapter examines the early integration of generative AI (GenAI), particularly large language models (LLMs) like ChatGPT, into judicial workflows. Unlike traditional rule-based decision-support systems, GenAI adopts a bottom-up approach, generating insights from vast datasets to assist real-time decision-making. While offering speed and improved access to information, these tools also present challenges that require careful understanding by their users. Using the recent case of a Dutch judge who employed ChatGPT to estimate the lifespan of solar panels, the chapter illustrates how GenAI is already being used in courtrooms. The value of GenAI lies in supporting, not replacing, human judgement. Yet without a clear grasp of how these systems work, including their limitations and potential biases, judges risk relying on opaque or flawed outputs. The ‘black box’ nature of LLMs complicates their responsible use and raises concerns about the balance between efficiency and discretion. The chapter argues that effective integration of GenAI depends not primarily on regulation, but on judicial education and critical awareness of the technology’s capacities and constraints.