Agenda item

To review the Artificial Intelligence Policy one year from its adoption at Cabinet. The Business Change and Intelligence Manager to provide members with a presentation at the meeting.

Consideration was given to the presentation of the Business Change and Intelligence Manger who asked the members to review the Artificial Intelligence Policy one year from its adoption at Cabinet.

The Business Change and Intelligence Manger highlighted that:

• An AI Working Group had been established to oversee initiatives, share best practices, and ensure responsible deployment. Governance was supported by the ICT Strategy Board, which reviewed business cases for new AI applications.
• Regular briefings, workshops, and staff communications had been delivered to promote awareness and encourage the adoption of AI.
  • AI tools were being used to align policies across the partnership and to perform impact assessments more efficiently.
  • AI had assisted with triaging complaints, which in turn reduced delays and improved accuracy.
  • AI supported officers with sourcing data and evidence.
  • AI transcription and summarisation tools were being used to assist with the production of draft minutes and analyse trends.

• The Transformation Team was leading the partnerships AI adoption, ICT ensured secure deployment and integration; and Data Protection teams monitored compliance with GDPR.
• Further exploration of Microsoft Copilot would enable further development of methodologies to quantify efficiency gains and potential financial savings.

Members considered the update and made the following comments:

• Members asked how secure AI was when drafting policies or processing sensitive information.
  • The Business Change and Intelligence Manager clarified that Copilot operated within the Microsoft 365 environment, which was as secure as SHDC emails and document systems. Public AI tools such as ChatGPT and Google Gemini were blocked for officers;
    • The Business Change and Intelligence Manager would confirm if this applied to Members.

• Members asked if Copilot could be automatically installed on members’ devices.
  • The Business Change and Intelligence Manager agreed this would be ideal and would check with ICT and would report back to the members if this could be arranged.

• Members questioned if it was safe to use ChatGPT on their Council iPads for tasks like risk assessments.
  • The Business Change and Intelligence Manager responded that members should avoid entering personal or sensitive data into ChatGPT. Copilot was recommended for work-related tasks as it was secure and integrated with Council systems.
    • The Business Change and Intelligence Manager would confirm the official position on ChatGPT access for members.

• Members discussed how accuracy was ensured when AI produced outputs such as risk assessments or social media posts.
  • The Business Change and Intelligence Manager explained that AI was an assistant, not an autonomous decision-maker. All outputs must be reviewed and signed off by a human officer with subject expertise. SHDC was not currently using AI for automated decision-making.

• Members suggested that if officers were required to check everything that AI produced, would this not duplicate work.
  • The Business Change and Intelligence Manager reiterated that AI would reduce background work and speed up processes, but human oversight would remain essential to maintain quality and compliance.

• Members asked how we ensured transparency when AI-generated content was published externally.
  • The Business Change and Intelligence Manager answered that disclaimers for AI-generated content were under consideration and would be implemented for any material published externally.

• It was suggested that members would like to receive training and a demonstration on AI tools like Copilot.
  • The Business Change and Intelligence Manager confirmed that ICT had delivered staff training sessions and would check whether similar training could be arranged for members.

• Members asked how AI use was being monitored across the partnership services.
  • The Business Change and Intelligence Manager confirmed that an AI Use Case Register was being maintained by the working group, which captured all areas where AI was currently deployed and ensured data protection impact assessments were completed where necessary.

AGREED:

That following the presentation of the Artificial Intelligence Policy the Panels comments would be noted.