Enhancing Preparedness of Public Services to Manage AI risk in Multilingual Communication
Start date
September 2024End date
August 2025Overview
Recent investigations have linked various cases of patient deaths and complications within the NHS to poor multilingual communication practices, as staff struggle to find effective solutions to communicate with members of the public with limited English proficiency. Despite risks of inadequate accuracy and security associated with machine translation (MT) tools such as Google Translate, reports and studies have shown they are commonly used in public services, and their risks are not being assessed. This concern is common to the use of other language AI tools, such as ChatGPT, in other public authorities.
This Arts & Humanities internally funded impact project is based on a study conducted by Dr Felix do Carmo and his team in 2023-2024, in which a Freedom of Information (FoI) request was sent to all NHS trusts, police forces and councils in England. This found numerous causes for concern regarding multilingual communication practices and a lack of preparedness to manage the risks associated with the use of MT and AI tools for an increasingly multilingual population.
This new project will see the team providing a meaningful contribution to the improvement of multilingual communication within public services, when using language AI tools, with a particular focus on the NHS, police forces and councils. They will design a training framework for the management of the risk of miscommunication through AI and pilot this with the NHS. In addition, they will organise similar workshops with police forces and local councils, applying the lessons learned from the NHS context and offering tailored training.
Team
Principal Investigator
Dr Felix do Carmo
Senior Lecturer in Translation and Natural Language Processing
Biography
I am a Senior Lecturer in Translation and Natural Language Processing, working on the application of natural language processing, machine learning, machine translation and assisted translation technologies in translation research, teaching and practice. Before joining the University of Surrey, I spent over 20 years as a translator, reviser, translation company owner, translator trainer, university lecturer and translation conference organiser in Porto, Portugal. In 2017, I moved to Ireland, to work for two years as a full-time post-doctoral researcher at the ADAPT Centre in Dublin City University, where I carried out research related to translation technology.
Co-Investigator
Dr Yuan Zou
Lecturer in Translation Studies
Biography
I am a Lecturer in Translation Studies at the University of Surrey's Centre for Translation Studies (CTS). My background spans audiovisual translation (AVT), interpreting, and post-editing. I hold a PhD in AVT from Queen's University Belfast (QUB) and an MTI in Translation and interpreting from Jilin University.
Before joining Surrey, I was teaching Interpreting and Translation at QUB, and I engaged in freelance work as a translator and interpreter. These experiences have been instrumental in shaping my research direction and pedagogical approach.
I am currently focused on the integration of language technologies in the fields of interpreting and audiovisual translation (AVT), with a keen interest in harnessing these advancements to improve digital accessibility. I am actively investigating innovative ways in which technology can be harnessed to support and improve access for individuals with disabilities, ensuring that digital content is more inclusive and accessible to all audiences.
Planned Impact
The team will produce four new reports that contain the results of the FoI study, related to 41 Police authorities and 314 councils in England. A register of identified risks and needs in the NHS and other public authorities will be created, along with training modules to support better management of MT for multilingual groups. Training outputs will include instruments for identification of available resources, tracking of decision-making procedures and impact evaluation. Together, these outputs will become part of an overall training framework. Longer term, the team will produce reports on factors that should be considered in designing policies for use of AI and MT in public services.
With the aim of seeing immediate changes in awareness, the project will not only highlight the risks associated with the use of MT and AI tools in communication in healthcare and other public services, but also lead to a greater acknowledgment of the role of human interpreters and translators in managing and reducing the risk of these situations. The implementation of the training programmes is expected to lead to clear changes, contributing to the development of best practices and increased reach at different levels, including local and nationwide public organisations. Evidence of these outcomes may be visible in policies and better financial management of contracts and procurement processes for human resources (translators and interpreters) and technologies.
Successful implementation of the project will improve communication processes involving linguistically diverse service users of the NHS, police forces and council services, with impact on both public service staff and service users. Staff in these authorities will feel protected by clear policies that will inform and guide their decisions regarding the use of translation and interpreting services with MT/AI communication tools, and clarify accountability within organisations, in the eventuality of problems occurring. Service users will benefit from improved, more accurate and safer communication support, resulting in better services (including improved patient safety in the NHS and fair access to public resources).
- Collect information on needs, risks and recommendations for training, regarding AI-related multilingual communication in the NHS
- Design of a modular CPD training programme with a set of thematic sessions for training different groups of NHS staff
- Programme piloted with 3 different groups of NHS staff
- Impact analysis report, which identifies factors that contribute to better multilingual communication policies
- Delivery of 1-2 impact workshops per target group
- Expansion of the components of the initial framework (1a. 1b.), adapted to facilitate its use across different public-sector organisations
- Delivery of pilot training to police forces and local councils
