Artificial intelligence in support of accessibility: The Brain-friend device

We live in a context where accessibility and inclusivity are key issues in the socio-technological landscape, ensuring equal access to information and resources is essential to allow each individual to fully participate in society and enjoy their rights. This concept not only concerns the removal of physical barriers, but also of digital and cognitive barriers, where people with visual impairments may have difficulty accessing unstructured online content for reading via screen readers, while those with hearing impairments may encounter obstacles with multimedia content without subtitles or transcripts, just as inclusiveness is not limited to physical disabilities, but also includes cultural, linguistic and cognitive differences. In this context, generative artificial intelligence offers tools and techniques to improve accessibility and promote inclusivity, thanks to its capacity for machine learning and adaptation to individual needs, providing highly personalised solutions, understanding and interpreting a wide range of data, including text, images and sound, and adapting the way information is presented according to users' preferences and needs (Fernandes et al., 2014). This paper presents a case study characterized by an accessibility system and a related ‘brain friend’ device, an expert system that is able to determine the accessibility profile of a user considering a series of parameters with the aim of generating the content likely requested by the user according to his or her accessibility profile, and also a device for the interaction with the user and the analysis of content according to the accessibility profile determined by the system. Brain friend represents an important tool for accessibility for digital cultural heritage applications. This paper shows also new possible approaches of the use of “brain-friend” to implement “custobot” experiences and also “human firewall” applications.