Institutions across higher education, legislative bodies, the judiciary, public broadcasting, and public health agencies exert significant influence through their communication practices, yet they often reproduce biases embedded in traditional linguistic conventions. Recent global events have also emphasized the imperative for these institutions to guarantee both the broad dissemination of information and the protection of public well-being via accessible, unbiased, representative, and inclusive messaging. Although the rapid expansion of large language models offers promising avenues for automating inclusive and accessible language use, these systems often mirror and even exacerbate existing prejudices. Addressing these challenges demands a transdisciplinary response: computer scientists and linguistics must collaborate to unpack the distinct and cross-linguistic gender markings and manifestations of bias, as well as the features of accessible language, all within a human-centered AI paradigm. By anchoring AI training in a multilingual ‘gold standard’ derived from rigorously curated linguistic resources, we can recalibrate and steer automated outputs toward unbiased language, gender sensitivity, and accessibility, thereby promoting equity in institutional discourse. 

The LINFA: Linguistic & Informatic insights on AI for gender and inclusivity workshop aims to translate our research foundation into concrete, collaborative action by convening scholars and practitioners from different disciplines to advance AI-driven strategies for equitable, gender-inclusive, and accessible institutional communication. We invite submissions that critically examine existing language technologies, evaluate or propose novel multilingual resources, explore the sociolinguistic dimensions of equitable, gender-sensitive, and accessible discourse across different languages, and showcase prototype applications or empirical assessments of user impact in real-world contexts.

Submission Guidelines

All papers must be original and not simultaneously submitted to another journal, conference, or workshop. The following paper categories are welcome:

• Short Papers: An extended abstract between 2 and 4 standard pages.

 All submissions should follow the provided template, available in both LaTeX and Microsoft Word formats.

List of Topics

•  gender bias and inequitable messaging in institutional communication

•  human-centered AI paradigms and bias-mitigation strategies

•  design and curation of multilingual resources for inclusive training data

•  cross-linguistic and sociolinguistic analyses of gender marking, bias, and/or prejudicial discourses

•  critical evaluation of LLMs and other language technologies for automated inclusive language

•  obstacles to comprehensibility and accessibility in institutional texts

•  plain-language and accessibility best practices for institutional communication

•  prototype applications of AI tools in real-world institutional contexts

•  empirical methodologies for assessing user reception, impact, comprehension, and/or trust

•  frameworks and metrics for evaluating and guiding AI-driven inclusive messaging

•  organizational, legal, and policy considerations for sustainable deployment of inclusive AI systems

Venue

The workshop will be held in Salerno, Italy in conjunction with CHItaly 2025.

Contact

All questions about submissions should be emailed to:

- Daniela Vellutino : dvelluitino@unisa.it

- Jacqueline Aiello: jaiello@unisa.it

- Mafalda Ingenito: mingenito@unisa.it