COMHAIRLE CONTAE ÁTHA CLIATH THEAS
SOUTH DUBLIN COUNTY COUNCIL

MEETING OF SOUTH DUBLIN COUNTY COUNCIL

Monday, September 08, 2025

QUESTION NO. 9

QUESTION: Councillor Y. Collins

Noting that the Council is open to using AI, how will it manage practical applications like e.g. traffic management, alongside concerns about transparency and public trust

REPLY:

The Council acknowledges your query regarding the management of AI applications, such as traffic management, in the public sector, particularly concerning transparency and public trust. The council is committed to an approach to AI that is ethical, accountable, and aligned with public service principles.

The practical application of AI is a complex and nuanced matter, as each use case is different. Our approach will be guided by the principles and regulations set forth in the EU AI Act, the government's Guidelines for the Responsible Use of AI in the Public Service, and any sectors guidelines that may be issued. The EU AI Act provides a legal framework to ensure that AI systems are safe and respect fundamental rights. It classifies AI systems based on their potential risk and establishes clear requirements for high-risk applications.

When deploying AI, our implementation will be based on seven core principles for trustworthy AI:

• Human agency and oversight: AI systems must be designed as a tool to support, not replace, human decision-making. Humans will retain the ability to understand, contest, and override automated decisions.
• Technical robustness and safety: AI systems must be resilient to errors and security threats. We will ensure systems are robust, reliable, and capable of operating safely under various conditions.
• Privacy and data governance: The council is committed to safeguarding personal data. We will adhere to strict data protection laws, ensuring that all data is handled with the utmost care, with a focus on privacy by design.
• Transparency: We will be transparent about the use of AI systems, providing clear information on their capabilities, limitations, and how they impact the public.
• Diversity, non-discrimination, and fairness: Our AI applications will be developed and used in a way that avoids biases and promotes equal access and fair outcomes for all citizens.
• Societal and environmental well-being: We will consider the broader impact of AI systems, aiming for solutions that contribute positively to society and the environment.
• Accountability: We will establish clear accountability frameworks for all AI systems, ensuring that there is a responsible party for every stage of the AI lifecycle, from design to deployment.

In addition to these principles, we can only consider AI solutions where our organisation's data remains entirely secure and is never used for training purposes. This is a crucial requirement to protect sensitive public information and maintain public trust. Any AI model or platform we adopt must be able to guarantee the complete isolation and security of our data, ensuring it remains within our control and is not utilised to improve or retrain the provider's general AI model. This commitment is paramount to our duty as a public service entity.