How to recognise algorithmic bias

13/01/2026

2 min.

Ethics and fairness in the use of artificial intelligence in the classroom: free webinar for teachers

How can algorithmic bias be recognised and countered? What considerations are needed for the fair and inclusive use of AI in teaching?

These questions will be addressed in the free webinar “Ethics and fairness in the use of AI in the classroom”, scheduled for Thursday 15 January at 5 p.m., promoted by the Fondazione Mondo Digitale ETS.

The meeting is designed to support teachers in the conscious integration of artificial intelligence into educational contexts, offering practical tools and food for thought on one of the most topical issues in the pedagogical debate.

During the webinar, the following topics will be addressed in particular:

how algorithmic biases arise and why it is important to be able to recognise them;
concrete examples and short simulations to understand the most frequent distortions;
a specific focus on gender bias and its impact on educational pathways;
teaching strategies and tools to mitigate bias and promote the fair, inclusive and responsible use of AI in schools.

The webinar will be led by Barbara Avella, a teacher and trainer with extensive experience in educational innovation and STEAM education, and is designed to offer content that can be immediately transferred into everyday classroom practice. At the end of the meeting, participants will be able to download a certificate of attendance and a series of worksheets to use with students.

The initiative is part of the Ital.IA Lab for School project, carried out by the Fondazione Mondo Digitale in collaboration with Microsoft, and works in synergy with the national programme Coding Girls & Women, supported by the Ministry of Labour and Social Policies under Notice 2/2024, reinforcing the joint commitment to reducing gender gaps and promoting informed digital citizenship.

The speaker

Barbara Avella has been a primary school teacher since 1985 and is a pioneer in STEAM teaching with an interdisciplinary approach aimed at overcoming the gender gap. She has been collaborating with the Fondazione Mondo Digitale for over ten years and is a Scientix Ambassador and ESERO Ambassador. She participated as a speaker at the MIT AI & Education Summit 2025, contributing to the discussion on the use of AI tutors in teaching. She is an author, science communicator and manager of highly interdisciplinary projects.

To participate

The card, download and spread the word

An example: staff selection

Research from the University of Washington analysed the use of artificial intelligence models in automated CV selection, highlighting significant gender, race and intersectional biases.

The study tested three large open-source language models (LLMs) on over 550 real CVs and more than 500 job advertisements, for a total of over 3 million comparisons. By varying only the names (associated with white and black men and women), the researchers observed that:

names perceived as white were favoured in 85% of cases, compared to 9% of names associated with black people;
names associated with men were preferred in 52% of cases, those associated with women in only 11%;
the systems never favoured black male names over white male names.

According to the researchers, these results demonstrate that the use of AI in recruitment processes, which is already widespread, can reproduce and amplify existing discrimination, making training, transparency and critical evaluation of these systems urgent, especially in educational and professional contexts.

[source AI tools show biases in ranking job applicants' names according to perceived race and gender]

How to recognise algorithmic bias