In a recent episode of the Science & Policy Exchange podcast, host Zahra Soar sat down with Rashid Mushkani, a PhD candidate in Interdisciplinary Planning at the University of Montreal and Mila, the Quebec AI Institute, to explore the critical intersection of urban planning, artificial intelligence (AI), and Equity, Diversity, and Inclusion (EDI). Their conversation delved into what it takes to build inclusive, accessible, and sustainable cities for all members of our communities.

What is Urban Planning?

For those unfamiliar with the term, Rashid Mushkani defines urban planning as the coordination of land use, services, and infrastructure to create people-centered, fluid, and accessible cities. It’s a balancing act, weighing social, economic, and environmental objectives within legal and political frameworks to translate a long-term vision for a city into day-to-day realities.

Crafting Inclusive Public Spaces

What does an inclusive public space truly look like? According to Mushkani, it’s a space that is personal, democratically managed, and addresses the diverse needs of the community. Inclusivity is reflected in both the physical elements of a space — like its materials and accessibility for individuals with disabilities — and in its governance.

The COVID-19 pandemic highlighted the essential value of these public spaces, shifting our collective understanding and prompting “tactical urbanism” — quick, on-the-spot solutions to community problems that demonstrated the possibility of rapid change.

However, what people consider “inclusive” can vastly vary from one person to another. For example, in a recent study published in the Journal of Urban Management, Mushkani and his colleagues found that when individuals were asked about their perceptions of Montreal streets, there was very little agreement on social criteria such as inclusivity and practicality. This indicates that perceptions of public spaces are deeply personal. Interestingly, individual assessments converged after small-group discussions, highlighting how collaborative dialogue and community engagement can help reconcile differing perspectives and guide the design of truly inclusive spaces that meet the needs of all community members.

Climate Change, Inequity, and the Urban Landscape

The conversation also touched on the pressing issue of climate change and its disproportionate impact on vulnerable populations. Mushkani points out that historically, social and economic inequities have shaped our cities.

This global trend is evident at the local level. In Montreal, for example, research has shown that populations of lower socioeconomic status have significantly less access to urban green space, which exposes them to greater risk from the urban heat island effect. These environmental injustices mean that lower-income and racialized communities often bear the brunt of climate risks.

In addition, Mushkani described how certain boulevards in Montreal have created a physical and visual barrier between affluent and poorer neighborhoods, a stark example of embedded inequality. For example, on Boulevard de l’Acadie, a green barrier separates more affluent neighborhoods from poorer ones, forming what some have dubbed the “Great Wall of Montreal”. This is further explored in his research in Habitat International, which shows how some Montreal streets were identified as “underperformers” after engaging individuals from various social positions in a “Street Review” to gauge their perceptions of the streets in terms of inclusivity, accessibility, aesthetics, and practicality. This research highlights how Boulevard de l’Acadie was often described by older adults and immigrants as very hostile, with limited cultural openness and minimal appealing features.

Furthermore, in cities like Montreal, significant accessibility problems persist, making it difficult for many residents to navigate their own communities. The result of an AI-based analysis of 45,000 street images of Montreal, showed that the vast majority of streets remain inaccessible to wheelchair users (Figure 1). This work, developed through collaboration with over 30 Montréal-based community organizations, paints a stark picture of urban exclusion.

Figure 1: Accessibility map of Montréal. Blue areas are accessible for wheelchair users, while orange areas are not accessible. Image from Rashid Mushkani.

The data is reinforced by the lived experience of residents. As wheelchair user Linda Gauthier stated to CBC News in 2022, the feeling is clear: “We are not welcome”. This powerful combination of data and personal testimony underscores how the design of our public spaces can systematically marginalize members of our community

However, there are positive examples of change. In Montreal, community-led initiatives to create more green spaces, such as gardens, are not only addressing climate issues but also strengthening community cohesion.

The Double-Edged Sword of AI in Urban Planning

A major focus of Mushkani’s research is the role of AI in shaping our future cities. He sees potential for AI to help design more inclusive public spaces, for instance, by using text-to-image models to help citizens visualize their ideas.

However, AI is not a silver bullet. Mushkani cautions that AI systems themselves have significant problems that need to be addressed. A primary issue is bias, which stems from two main sources:

• Data: AI models are often trained on data that reflects existing societal biases. If historical data shows that certain groups have received fewer resources, an AI trained on this data will learn to perpetuate, and even amplify, that inequity.
• Methods: The training process itself can reinforce a single, dominant worldview, overlooking the diverse perspectives and truths essential for genuine equity.

This bias can manifest in service allocation, where AI might inadvertently direct more resources to already privileged neighborhoods. The failed Sidewalk Toronto project serves as a cautionary tale. The smart city initiative by a Google-affiliated company was cancelled in May 2020, officially due to economic uncertainty, but only after years of significant public opposition centered on concerns over data privacy, surveillance, and a lack of transparent governance. The first question, Mushkani argues, should always be: “Do we really need it?”.

The Path Forward: Citizen Participation and Policy

So, how do we move toward more equitable cities? Mushkani repeatedly emphasizes one core solution: participation.

He argues that for cities to be truly equitable, development and decision-making must be citizen-driven. Research and practice have shown that robust citizen engagement can lead to better development outcomes, including improved service delivery, greater community trust, and a more equitable allocation of public resources. When citizens are aware of their rights and actively participate in the planning process, they can hold tech companies and governments accountable, pushing for transparency and legislation that serves the community’s interests.

For students and professionals inspired to work at this intersection, Mushkani offers this advice: work with community centers, engage with minority groups, and adopt a transdisciplinary approach to research and problem-solving.

This podcast is brought to you by the team of Science and Policy Exchange — learn more at https://www.sp-exchange.ca/.

We sincerely thank Rashid Mushkani for this valuable discussion and we acknowledge the contributions of the team that made this podcast episode possible:

• Interview by Zahra Soar
• Script writing by Zahra Soar, with the support of Cassandra Johannessen, Gloria Lau, and Jonathan A. Caballero
• Audio Editing by Jacob Van Oorschot
• Executive production by Jonathan Caballero

If you have any comments or would like to share with us topics that you would like to learn about in the podcast, please reach us at contact@sp-exchange.ca.