The early stages of a proposed plan to transform part of Hamilton's industrial waterfront into an artificial intelligence data centre encountered a significant setback last week. Toronto-based Slate Asset Management, which acquired approximately 800 acres of industrial land and buildings from Stelco four years ago, announced its intention to redevelop the site into a world-class industrial park focused on job creation, economic growth, and environmental sustainability.
Steelport Project's Potential Impact
The area, known as Steelport, has the potential to generate up to 23,000 new jobs across the Greater Toronto and Hamilton Area and inject up to $3.8 billion into Ontario's economy, according to the company. However, local residents opposed to the early plans, which would involve subdividing part of the land, gathered outside Hamilton City Hall on Thursday morning to protest ahead of a committee of adjustment meeting considering the owner's application.
Tribunal Decision
Tribunal member Robert Reid acknowledged that the application for land severance met all regulatory requirements for the property. Nevertheless, after hearing numerous residents argue against the application in protest of AI infrastructure, along with thousands of emails and messages submitted to the committee, the tribunal voted 3-0 to deny the application. Residents at the meeting erupted into loud cheers upon hearing the decision.
Despite this denial, the developer has the option to appeal the decision to the Ontario Land Tribunal or proceed with its plans without subdividing the land.
Broader Skepticism Toward AI
The protest occurs amid growing skepticism among Canadians about AI, which could replace jobs while consuming significant amounts of water and power required to run energy-intensive AI processors.
What Is an AI Data Centre?
An AI data centre is a large specialized facility that houses the technological infrastructure needed to train, deploy, and deliver AI applications and services. According to CaTECH Systems, a telecommunications contractor in Markham, these supersized data centres require substantial energy to operate hardware and software, as well as massive amounts of water for cooling systems. The company notes that as AI continues to disrupt industries, the demand for specialized AI data centres has grown, with these next-generation facilities custom-built to support the heavy and complex needs of AI workloads. In contrast, traditional data centres are designed for predictable, incremental, and less power-intensive computer-based workloads.
Researchers at York University report that there are currently five AI data centres operating in Canada: two in British Columbia, and one each in Ontario, Quebec, and New Brunswick.
Controversies Surrounding AI Data Centres
Critics cite energy and water consumption, land use, and noise pollution as major issues. The Starfish Environmental Society, based in Vancouver, highlights that large amounts of electricity are needed to run these facilities, with many relying on fossil fuels. Furthermore, many data centres install backup natural gas or diesel generators to ensure uninterrupted power supply, contributing to increased greenhouse gas emissions. Additionally, AI data centres consume large amounts of fresh water for cooling, and there are complaints about noise pollution from the constant drone of the facilities and the extensive land required to house the technology.
Canadian Public Opinion on AI
According to a recent Angus Reid survey, more than two-thirds (68%) of Canadians want governments to heavily regulate AI, and nearly three-quarters (74%) doubt any government can keep pace with the growing technology. Moreover, 68% said they would oppose a large AI data centre near their home, while 79% support taxing companies that replace workers with AI. The pollster notes that Canadians see some upside to data centres but grow more skeptical as the impacts move closer to daily life. The only positive result from the survey found that 38% of respondents think Canada can compete globally in AI and tech, while 20% disagreed, and a plurality (42%) were unsure.
