Last week’s webinar “The Energy Transition Is Global – Reflecting A Newfound Respect for the Environment” can be viewed HERE   https://youtu.be/RcpTC7GL42M featuring an interview of Leigh Bond and discussion with his team. Also consider Michael Campbell’s interview HERE of Dr. Judith Currie (minutes 6 through 27) acknowledging climate change but uncertain of the risks: that the warming is an urgent existential crisis, that weather events reflect climate change, and challenges the anti-capitalist activism mandating of extraordinary measures such as Canada’s Carbon Tax.

---

AI in Municipal Governance: Opportunities and a Few Cautionary Tales

Municipalities are increasingly using AI for various purposes, including citizen services, urban planning, public safety, administrative tasks, data analysis, public benefits, infrastructure management, and healthcare..The future of AI in municipal governance is promising but requires balancing opportunities with risks and ensuring responsible and ethical use of the technology.

Jeff Uhlich BA, MSc HRM, is a retired human resources executive and consultant in Edmonton, who writes on the intersection of talent management, technology and governance.

A few weeks ago, Perry shared an article with some of his contacts about Nevada’s implementation of AI to help manage the State’s unemployment claims. The COVID-19 pandemic had left the state struggling with a backlog of 40,000 claims, and they turned to AI for a solution. As reported in the article, the State’s Department of Employment, Training, and Rehabilitation (DETR) is planning to use an AI system developed in collaboration with Google to process unemployment benefit appeals. This system will analyze hearing transcripts and evidence to make recommendations, with human “referees” reviewing the AI’s decisions before they are finalized.

‘In as little as five minutes, the AI will issue a ruling that would have taken a state employee about three hours to reach…’.

But the system has been controversial, with legal and labour experts raising concerns about potential errors and the risk of bias. There are also worries about the pressure on human reviewers to approve AI recommendations quickly and whether courts will be able to review or overturn AI-influenced decisions because current Generative AI models can be opaque about how they produce their response. This lack of observability and transparency is problematic and there’s been a lot of pushback from citizens who are expressing concern about abdicating human responsibility to an AI. The 173 comments posted on the article Perry shared convey this concern in very frank terms.

The Nevada example piqued my curiosity and got me exploring the bigger questions about how and where municipalities are implementing AI solutions to improve their efficiency and effectiveness. What I learned is that there’s a quiet revolution happening in municipal governance. From urban planning to public safety, municipalities are harnessing the power of AI, machine learning, natural language processing, and big data to transform the way they operate. These pioneering municipalities offer lessons to us all.

When you’ve looked at enough examples of how municipalities around the world are implementing these tools you’ll see that they fall into several categories and that there are varying levels of success in how the use of these tools is being received. There are eight big buckets:

1. Citizen Services and Engagement

• Example: Singapore’s “Ask Jamie“, AI-powered chatbots for constituent support
• Generally well-received due to improved accessibility and response times
• Less controversial as these tools primarily provide information rather than make decisions

1. Urban Planning and Resource Management

• Example: Barcelona, Spain – Digital City Initiative: Barcelona uses AI and IoT sensors for urban planning, including managing traffic, waste collection, and energy usage.
• These types of initiatives are often successful due to data-driven insights improving efficiency
• Generally, less resistance as they support human decision-making rather than replace it

1. Public Safety and Law Enforcement

• Examples: PredPol (discontinued), Axon’s Draft One (disclosure: I own shares in Axon Enterprise)
• Mixed reception: efficiency gains are appreciated, but there are major concerns about bias and privacy
• There’s higher scrutiny and potential pushback on these uses due to the sensitive nature of the application

1. Administrative and Internal Operations

• Examples: AI for financial narrative writing, policy draft generation, fraud detection
• Generally positive reception for improving efficiency and reducing mundane tasks
• These uses show less public resistance as they primarily affect internal processes

1. Data Analysis and Decision Support

• Example: AI systems analyzing urban trends, demographic shifts
• These use cases are often successful in providing valuable insights to human decision-makers
• They are less controversial when used as a support rather than for autonomous decision-making

1. Public Benefits and Social Services

• Examples: Nevada’s unemployment claims system (the article linked above that started this all and this one)
• Mixed reception: there’s a high potential for efficiency, but also high scrutiny due to direct impact on citizens
• Experience shows significant concern about fairness, transparency, and the ability to appeal decisions

1. Infrastructure and Environmental Management

• Examples: AI for smart cities, traffic management, environmental monitoring
• Generally positive reception due to potential for improved urban living and sustainability
• There’s less resistance when efforts are focused on optimization rather than enforcement

1. Healthcare and Public Health

• Examples: AI in local healthcare systems, public health monitoring
• Potential for significant benefits, but privacy concerns may lead to pushback
• Success often depends on clear communication of benefits and robust data protection

Under the broad label of ‘AI’ there are many subcategories — it’s not a monolith. What people generally describe as AI today refers to Generative AI because of the hype around ChatGPT. But from the examples above it’s obvious that Generative AI is just the tip of an iceberg that’s been building for decades. In the area of municipal governance, machine learning, natural language processing, and big data, are as, if not more, important than Chatbots.

For municipalities implementing these tools, success seems to aggregate in the following areas:

• Data analysis and decision support tools
• Citizen information services
• Infrastructure and resource optimization
• Internal administrative processes

Conversely, there are areas of use by municipalities that are facing real resistance and scrutiny:

• Autonomous decision-making systems, especially in sensitive areas like law enforcement or benefit determinations
• Applications involving personal data that could impact individual rights
• Systems perceived as replacing human judgment rather than supporting it

What differentiates these use cases? What have these pioneering municipalities taught us about how and where to employee these tools? The key factors appear to be:

1. Transparency in how AI is used
2. Maintaining human oversight and decision-making power
3. Clear benefits to efficiency or quality of service
4. Robust safeguards for privacy and fairness
5. Focus on augmenting rather than replacing human capabilities

The benefits of using these tools are evident. Municipalities managing these factors can increase their efficiency, improve decision-making, enhance citizen engagement, proactively manage public safety, and save money through process automation and optimization.

But the risks must be understood and managed. Bias, transparency, privacy and security, job displacement, public trust and acceptance, and error correction are issues that pioneers have struggled to solve.

As more municipalities adopt AI, it’s important to address these risks proactively and ensure that the benefits of AI are realized in an ethical and responsible manner. Organizations considering future AI adoption can start by examining their people, processes, and technologies.

• Are your people ready or do they need training and development to harness this opportunity? If you don’t have the staffing resources or budget for training, do you have it for consulting?
• Are your processes ready? Are your processes well understood and documented? Are they ‘clean’? In other words – don’t automate a ‘bad’ process.
• And finally, is your technology stack ready and are your privacy and data security systems and policies up to date and ready for the implications of AI adoption?

Municipalities need to assess the form, quality, and sources of their mission-critical data, determine what percentage is digital versus paper-based, and develop a plan to make this data machine-readable and accessible to AI systems. Then, consider this information in light of the people, process, and technology priorities I’ve described. Luckily, the new crop of Generative AI tools is now multi-modal which means they can much better process images, so ingesting text from a scanned document or even a photograph is possible and improving rapidly.

The age of AI in municipal governance is upon us and the examples set by these pioneering municipalities will pave the way for others. While larger, more technologically sophisticated municipalities are leading the charge, there are valuable lessons for smaller entities and even small businesses. We all need to stay informed and find that balance between embracing the opportunities and mitigating the risks. We can start small, invest in areas with a higher probability of success and avoid unnecessary risks. If we do that, we can shift the future of governance and service delivery in a way that harnesses the power of AI for the benefit of our communities and ensures that this technology is used effectively, responsibly and ethically.

---

https://us02web.zoom.us/j/84258596166?pw..

https://us02web.zoom.us/j/84258596166?pw..

---

Editor@KEInetwork.net

Visit KEInetwork.net