Dangers of co-thinking with GenAI: The biases and ethical implications

Generative Artificial Intelligence (GenAI) is increasingly embedded in innovative learning design, and its role is shifting towards a collaborative thinking partner that participates in learners’ reasoning, framing, and judgement. This chapter examines the dangers that emerge from this shift and its implications on innovation learning labs, where experimentation and problem-solving foregrounds learning and reflection.
Drawing on a critical synthesis of academic literature and contemporary discourse on GenAI, this chapter identifies two recurring patterns relevant to innovation-focused pedagogy. First, it outlines how bias is amplified through pedagogical normalisation, as repeated co-thinking with GenAI stabilises particular modes of reasoning, language, and problem framing that can narrow the exploratory space central to innovation learning. Second, it examines ethical slippage in everyday co-thinking, where responsibility, authorship, and evaluative judgement become increasingly diffuse through routine interaction with GenAI, even in the absence of intentional misuse.

The discussion interprets these patterns through the aims of the Innovation Learning Lab. It argues that GenAI’s capacity to accelerate thinking risks eroding the frictional moments of uncertainty, resistance, and struggle that are essential for transformative learning. In response, this chapter frames ethical learning as a situated practice that must be deliberately designed into innovation learning environments. By situating biases and ethical risks within the pedagogical conditions of the Innovation Learning Lab, the chapter contributes a critical lens for engaging GenAI in innovation education without undermining the cognitive, ethical, and reflective capacities that such environments seek to cultivate.

keywords: generative artificial intelligence; innovation learning labs; ethical learning

• with: Floris Van der Marel
• year: November 2025 — ongoing