As part of the DIVERSIFAIR Erasmus+ project, CorTexter has developed an AI-supported language tool designed to assist organisations in producing clearer, more inclusive job advertisements. This tool contributes to the project’s mission to advance intersectional fairness in and for AI systems.
Key features
- Bias detection: The tool scans job postings for five common types of linguistic bias and highlights areas that may unintentionally discourage certain applicant groups.
- Inclusive language suggestions: It identifies jargon and vague terms and offers alternative phrasing that is more accessible and engaging for diverse audiences.
- Multilingual support: Available in British and American English, Dutch, German, and French, enabling organisations to reach a broader range of candidates in their native languages.
- Intersectional fairness focus: Developed to support the reduction of bias across multiple identity dimensions, reflecting DIVERSIFAIR’s commitment to intersectionality in AI ethics and fairness.
The CorTexter tool was created as a research output within the DIVERSIFAIR project, aiming to provide practical resources for organisations to improve recruitment practices. his tool is offered as a means to critically engage with the promises and limitations of AI-driven language interventions. It provides a space for testing and reflecting on how such systems may support – but also risk undermining – efforts toward more inclusive hiring practices, particularly when evaluated through an intersectional lens.
