CHIWORK paper: Generative AI should support, not replace, human-centered design

In a new position paper for the CHIWORK '26 workshop (Interrogating GenAI Augmentation for CHIworkers), author Amir Talakoob examines how generative AI is permanently altering human-centered design. The paper, posted on arXiv (2608.07483), argues that the AI wave accelerated by ChatGPT's public launch in November 2022 has introduced lasting changes in how designers and researchers understand users, generate ideas, evaluate systems, and make design decisions. While some adoption reflects hype and overuse, Talakoob contends many shifts will outlast the initial enthusiasm—making it critical to examine what is gained and lost when AI enters the design process.

The paper identifies concrete risks: over-reliance on AI tools that erode human judgment, algorithmic bias embedded in generated insights, data security concerns from feeding user research into AI systems, and unclear accountability when AI-informed decisions go wrong. To preserve the human-centered core of design, Talakoob calls for deliberate AI literacy training, careful ethical judgment, and clearer boundaries defining where AI assists versus where human expertise must lead. The argument positions AI as a support mechanism—not a replacement for the designer's interpretive, empathetic, and ethical role. As GenAI increasingly automates research synthesis and ideation, this work pushes back against uncritical adoption and asks practitioners to define intentional guardrails before those boundaries become set by default.