As artificial intelligence becomes a common tool in education, its influence on students' perceptions of professional identity must be critically examined. This poster presents a corpus-based study analysing AI-generated English job descriptions for gender-neutral professions. The corpus comprises over 600 job descriptions (approx. 350,000 words) across fields such as STEM, education, public administration, and healthcare. Half of the texts were generated using a leading large language model, while the remaining half were sourced from real-world online job advertisements to provide a comparative baseline.

Despite English being considered a gender-neutral language, our analysis reveals that AI language models often reproduce traditional gender stereotypes—associating agentic qualities (e.g., leadership, assertiveness) with roles typically perceived as male-dominated, and communal traits (e.g., empathy, cooperation) with roles stereotypically linked to women.

The educational implications of these findings are significant. When students interact with AI-generated content—for example, during career counselling sessions, language learning tasks, or mock application writing exercises—they may be subtly influenced by the gendered framing of professional roles. Such exposure can reinforce internalised stereotypes, affect students’ career choices, and perpetuate existing gender imbalances in fields like engineering or caregiving. Furthermore, educators who rely on AI tools for generating teaching materials or assignments may unintentionally pass on these biases.

This study calls for a proactive integration of critical digital literacy into curricula, encouraging both teachers and students to question and deconstruct AI-generated language. Strategies for inclusive language education addressing algorithmic bias are proposed. They include comparative text analysis exercises, classroom discussions on representation, and AI bias awareness modules. By equipping students with the tools to critically engage with AI-mediated texts, educators can help foster more equitable and reflective learning environments.

Keywords: Gender stereotypes, language bias, artificial intelligence, educational equity.