Canadians have valid concerns about AI and jobs, privacy, fraud and the environment

Canadian research points to real risks, while evidence on employment and resource use calls for careful interpretation

THUNDER BAY — October 14, 2026 — Canadians who worry about artificial intelligence have valid reasons to ask hard questions. Research and public warnings identify risks involving changing jobs, personal information, misleading content, fraud and the resources needed to power the technology.

The evidence is less certain about how far those impacts will reach. A job that AI can change is not necessarily a job that will disappear.

What is worth considering is that AI is just another new tool.

In Northwestern Ontario over the past four generations, in forestry it has shifted from a logger with a bucksaw, axe and pickaroon along with a horse, to a tree cutter with a chainsaw and a Tree Farmer pulling out the logs to today’s mechanical tree harvester.

All driven by industry change.

In the news business, back in the day type was set by picking out each lead letter and putting the newspaper page together. Starting in the 1950s that began to change, and at newspapers like the Winnipeg Tribune, workers were on strike protesting the new evolution of technology.

Photography in the newsroom has shifted. Film has been replaced by digital imagery. That has allowed the speed of reporting, especially online to move faster than ever.

Today newspapers and news gathering is increasingly electronic. Even the physical newspaper is starting to see diminished print runs.

All of this was because new tools came together.

Environmental costs also vary with the technology, its location and how it is powered.

Much of the current debate involves generative AI: systems that produce text, images, audio, video or computer code. Their growing abilities create opportunities, but also make oversight and accountability more important.

Job concerns are justified, but exposure does not mean unemployment

A May 2025 study by the International Labour Organization and Poland’s National Research Institute, NASK, estimated that one-quarter of global employment was in occupations potentially exposed to generative AI.

Clerical work had the highest exposure, with media, software and finance occupations also affected. The researchers concluded that changes to jobs were more likely than complete replacement, because many occupations still contain tasks requiring human involvement. The study assessed technological exposure, rather than counting future layoffs.

Canadian evidence also calls for caution. A Statistics Canada study released January 28, 2026, found that employment generally grew between November 2022 and December 2025 across groups with different levels of potential AI exposure.

However, growth in coding occupations was concentrated among workers aged 30 to 49. Employment among coding professionals younger than 30 stagnated.

The researchers could not establish whether these patterns reflected AI, economic conditions, demographic changes or several factors together. Their findings do not establish that workers are protected from future disruption. For workers, that distinction matters. Concern about changing career opportunities is reasonable. Predicting a particular person’s job loss from an occupational exposure score goes beyond what these studies can establish.

Training gaps add to the uncertainty

Statistics Canada’s analysis of business AI use in the second quarter of 2026 found that 32 per cent of businesses using AI reported training existing employees because of that use.

Among AI-using businesses with at least 100 employees, the figure was 68.1 per cent. Among those with one to four employees, it was 24 per cent.

These are business-level figures. They do not reveal how many individual workers received training or whether that training was adequate.

Analysis: The difference suggests that workers’ access to preparation can depend heavily on their employer’s size. Practical responses include paid learning time, clear instructions about approved tools and involving employees in decisions about how their work will change.

Confident answers can still be wrong

The U.S. National Institute of Standards and Technology identifies confidently presented false information as a core generative AI risk. This is often called a “hallucination.”

An answer can be polished and convincing while containing incorrect facts or invented references. NIST also identifies overreliance on AI as a risk: users can give its output more weight than it deserves.

That makes checking important claims essential. A useful draft or explanation should still be assessed against reliable sources, especially when an error could affect someone’s health, finances, legal position or safety.

Bias creates another concern. Canadian privacy regulators warn against AI uses that lead to unfair or discriminatory treatment. Their guidance calls for organizations to explain significant AI-supported decisions, the safeguards involved and the options available to affected people.

Personal information can become part of the system

Canada’s Office of the Privacy Commissioner warns that information entered into AI chatbots may be collected and stored. Depending on the service, it may also be used to train the system or create profiles.

The commissioner advises users to limit personal information, avoid sensitive or identifying details, review privacy settings and avoid sharing other people’s information.

For workplaces, that makes tool selection and data rules important. Before uploading customer records, personnel files or confidential documents, employees need to know whether the service is approved and how it handles that information.

A service’s privacy protections should be checked directly. Different products and account arrangements can have different policies.

Thunder Bay police have warned about AI scam videos

The local implications already extend beyond the workplace.

On August 28, 2026, Thunder Bay Police Service warned residents about AI-generated videos promoting investment opportunities, cryptocurrency purchases and quick-profit schemes.

Police said fraudsters use fabricated presenters and altered public figures to make offers appear legitimate. TBPS said it had received reports concerning this type of scam.

Its advice is to verify offers by contacting the company through a trusted source. Anyone who loses money or discloses personal or financial information should contact their financial institution immediately and report the incident to police and the Canadian Anti-Fraud Centre.

AI’s environmental footprint is real and uneven

AI relies on physical infrastructure, including servers, electricity supplies and cooling systems.

The International Energy Agency’s 2026 analysis projects that worldwide data-centre electricity consumption will roughly double from 485 terawatt-hours in 2025 to 950 terawatt-hours in 2030. That would represent about three per cent of global electricity demand.

Those totals cover all data centres, including services beyond AI. Electricity use by AI-focused data centres is projected to triple over the same period. These are forecasts, with uncertainty around adoption, efficiency and infrastructure constraints.

The IEA also reports major improvements in energy efficiency per AI task. However, applications such as video generation and complex reasoning can require substantially more energy than simple text generation. Electricity-price effects depend on local supply, infrastructure investment and policy choices.

Water is another consideration. A December 2024 Lawrence Berkeley National Laboratory report estimated that U.S. data centres directly consumed about 66 billion litres of water in 2023.

That estimate covers all U.S. data centres, not AI alone, and is not a Canadian total. The report also describes water consumption associated with producing electricity. Environmental impacts vary with location and the power sources supplying a facility.

These figures support asking developers about electricity demand, cooling and water consumption. They do not establish a uniform footprint for every chatbot request or a specific impact on Thunder Bay’s water supply.

Benefits need to be demonstrated alongside safeguards

AI can deliver measurable workplace benefits. A study published in the May 2025 issue of The Quarterly Journal of Economics examined 5,172 customer-support agents.

Access to an AI assistant increased issues resolved per hour by 15 per cent on average, with larger benefits for less experienced and lower-skilled agents. Results from that workplace do not establish the same gain for every occupation or employer.

Analysis: For Thunder Bay and Northwestern Ontario employers, a practical approach is to test whether a tool improves a defined task, account for the time needed to check its output, protect confidential information and give workers appropriate training.

Residents and employees should also be able to ask who remains responsible when an AI-supported decision goes wrong. Clear answers about responsibility, data use and resource demand are part of earning public trust.