EOITC AI Workshop - Nov 27, 2025
On November 27, 2025, the Eastern Ontario Information Technology Collaborative (EOITC) hosted a webinar on artificial intelligence (AI).
The goal was to help municipalities understand what AI is, how it’s being used, what benefits it might offer, and what risks or challenges municipalities need to think about.
If you weren’t able to attend, you can watch the full recording below. You’ll also find an AI-generated text summary and the slide deck we used, further down the page.
What This Webinar Was About
AI tools are becoming more common in workplaces, including the public sector. Many municipal staff are hearing about AI but aren’t sure:
What it can do
Whether it’s safe to use
How it could help their day-to-day work
What rules or policies should guide its use
This session brought together IT leaders and municipal staff to learn the basics, share early experiences, and discuss what responsible AI use might look like in a municipal environment.
Key Points from the Session
AI is already being used in many municipalities, mostly for small, practical tasks like drafting documents, summarizing long reports, or helping staff look up information faster.
Having clear rules is essential. Staff need guidance on what tools are allowed, what data can be used, and how to avoid privacy and security risks.
Skills and comfort levels vary. Some staff are keen to use AI; others are unsure or concerned. Training and support will be important.
Practical “small wins” are the best place to start, not major system changes.
Municipalities want to collaborate by sharing templates, use cases, and lessons learned so no one has to figure everything out on their own.
Meeting Recap
Generated by AI. Make sure to check for accuracy. 😁
Meeting notes:
King Township's AI Implementation Journey: Marco from King Township detailed their multi-year journey implementing AI, highlighting the evolution from initial exploration and low-risk pilots to a broad, staff-driven innovation model, with support from leadership and cross-departmental partnerships.
Initial Exploration and Strategy: Marco described how King Township began its AI journey in 2024 by forming a digital transformation framework, intentionally positioning AI as a tool to address council priorities rather than as an end in itself. Early efforts focused on aligning AI initiatives with strategic objectives, such as reducing information-only calls to the contact centre.
Low-Risk Pilots and Early Wins: The team launched Kingsley, an AI-powered virtual assistant, as a low-risk, high-impact pilot. Kingsley was grounded in public documentation, not connected to business systems, and provided after-hours support to residents. The tool saw high engagement, positive feedback from accessibility groups, and demonstrated value through usage data.
Expansion to Cross-Departmental Use Cases: King Township expanded AI applications to include Rover, a smart vision tool for pothole detection, and began integrating data from regional partners. The team also developed AI agents for invoice processing, policy search, report review, and certificate management, often in response to staff-identified needs.
Staff Training and Skills Development: Marco emphasised the importance of training staff in prompting and AI tool usage, noting that initial gaps in digital literacy required tailored, hands-on training. Over time, the focus shifted from basic prompting to enabling staff to build their own AI agents.
Leadership Support and Organisational Change: A leadership mandate tied AI and automation adoption to performance reviews for supervisors and managers, catalysing a surge in grassroots innovation. Marco highlighted the necessity of leadership buy-in and cross-functional partnerships, including clerks, communications, and strategy, to manage privacy, change, and messaging.
Survey of AI Adoption in Eastern Ontario Municipalities: Sumaya presented survey findings from 256 respondents across Eastern Ontario municipalities, revealing widespread AI adoption for both work and personal use, with most staff using tools like ChatGPT and Copilot for writing, summarisation, and workflow support.
Survey Scope and Respondent Profile: The survey, distributed to EOITC member municipalities, received 256 responses from a broad range of roles, including supervisors, clerks, finance, IT, and planning, ensuring diverse perspectives on AI adoption.
Current State of AI Adoption: A significant portion of respondents indicated their municipalities are in exploratory or pilot phases with AI, while a smaller group reported strategic or fully integrated use. Many staff are personally using AI for writing, summarisation, research, and workflow support.
Personal and Organisational Use Cases: Staff reported using AI for a variety of tasks, such as drafting emails, summarising documents, generating templates, and even personal tasks like travel planning and speech writing. Organisational use mirrored these trends, with additional applications in mapping, GIS, asset management, and public engagement.
Barriers and Concerns: Respondents identified privacy and data security as primary concerns, alongside worries about accuracy, workforce disruption, ethical and legal risks, and a lack of formal training or guidelines. Most staff expressed interest in receiving more training and having clear policies.
Opportunities for Collaboration and Training: The survey highlighted opportunities for regional collaboration, with some municipalities already partnering on AI initiatives. There is strong interest in developing shared policies, guidelines, and training programmes to support safe and effective AI adoption.
Risks, Security, and Governance in Municipal AI Use: Frank provided an in-depth overview of AI-related risks, including data breaches, hallucinations, vendor management, and prompt injection attacks, and emphasised the need for robust policies, staff training, and careful vendor assessment to mitigate these risks.
AI Hallucinations and Data Accuracy: Frank explained that AI tools, especially chatbots, can generate inaccurate or fabricated information ('hallucinations'), sometimes with significant consequences, as illustrated by legal and customer service incidents. He advised verifying AI outputs and being cautious with links and references provided by AI.
Data Privacy and Unintentional Disclosure: Examples were provided of staff and contractors inadvertently uploading sensitive or proprietary data to AI platforms, leading to potential data breaches. Frank demonstrated how AI models can be manipulated to reveal uploaded data, underscoring the importance of restricting what is shared with AI systems.
Prompt Injection and Security Vulnerabilities: Frank showcased prompt injection attacks, where adversarial prompts can bypass chatbot restrictions and extract confidential information. He stressed the need to limit AI access to sensitive data and to regularly test AI systems for such vulnerabilities.
Vendor and Application Risk Management: Municipalities were advised to scrutinise vendors for AI integration, data residency, and model training practices. Frank recommended including AI-specific questions in RFPs and ensuring vendors notify clients of significant model updates or changes.
Policy, Training, and Organisational Readiness: Frank and Ben emphasised the necessity of comprehensive AI policies, regular staff training on AI risks and safe usage, and ongoing awareness of how AI features are embedded in both new and existing applications. They advocated for a multidisciplinary approach to governance and risk management.
Municipal AI Use Cases and Best Practises: Ben and Marco shared a wide range of AI use cases from Canadian and international municipalities, including chatbots, document summarisation, asset management, accessibility, and predictive analytics, illustrating the breadth of opportunities and the importance of starting with low-risk, high-value projects.
Service Delivery and Accessibility: Municipalities are deploying AI-powered chatbots for resident inquiries, meeting summaries, and council report generation, with notable benefits in after-hours service and accessibility for users with disabilities.
Infrastructure and Asset Management: AI tools are being used for infrastructure monitoring, such as pothole detection via vehicle-mounted cameras, automated licence plate recognition, and predictive maintenance, enabling more efficient resource allocation.
Planning, Compliance, and Workflow Automation: Applications include AI-driven building code compliance checks, automated document redaction, policy search, and workflow automation for tasks like invoice processing and report review, reducing manual effort and improving consistency.
Public Engagement and Communication: AI is facilitating public engagement through sentiment analysis, translation services, and the simplification of complex documents, making information more accessible and fostering greater transparency.
International and Cross-Jurisdictional Examples: Ben highlighted examples from the UK and other regions, such as Swindon's document simplification service and San Jose's AI guidelines, demonstrating the global nature of municipal AI innovation and the value of learning from diverse contexts.
Developing an AI Framework and Next Steps: Ben outlined the collaborative process underway to develop an AI framework and policy for Eastern Ontario municipalities, emphasising the importance of principles, guardrails, and ongoing engagement to ensure responsible and effective AI adoption.
Framework Development Process: The Eastern Ontario IT Collaborative, with support from Perigroup Consulting, is leading a series of workshops to identify use cases, establish guiding principles, and draft a regional AI framework and policy, drawing on survey data and municipal experiences.
Principles and Guardrails: Emergent principles include fairness, human oversight, transparency, data protection, and piloting low-risk projects before scaling. Reference materials from King Township, the Canadian federal government, and international cities are being used to inform the framework.
Opportunities for Participation: Attendees were invited to participate in the ongoing framework development by signing up for future workshops and contributing to policy discussions, ensuring broad input and regional alignment.