1. The Quiet Shift in How We Know What We Know
There is something unsettling about the way trust has slowly dissolved in our everyday lives. You scroll through social media, and you see a video of a politician saying something shocking, a celebrity endorsing a strange product, or a community leader making a controversial remark. Your first instinct might be to share it, to comment, to feel angry or amused. But then a doubt creeps in: Is this real? In recent years, that doubt has grown into a constant companion, and the CityNews Montreal piece on artificial intelligence fanned misinformation online captures exactly why. The issue is not just fake news in the old-fashioned sense, but the enormous, terrifying power of AI to generate content that looks, sounds, and feels human. What was once the job of propagandists, tabloids, or attention-seekers has now become something anyone can do with the click of a mouse. An algorithm can produce thousands of articles, videos, and voice recordings in seconds, all designed to influence how we think, vote, buy, and even whom we love or fear. The scariest part is not that this technology exists, but that it has become woven into the fabric of our information environment so quietly that many of us have already absorbed its falsehoods without realizing it. We are not dealing with obvious errors anymore; we are dealing with polished, persuasive, and emotionally charged content that is engineered to bypass our critical thinking. The human brain, wired for stories and connection, is no match for a machine that has learned our biases and can exploit them at scale. This new reality is forcing us to redefine what “knowing” even means. In Montreal, as in every other city, people are waking up to the fact that every image, every voice, and every news headline must now be examined with a level of suspicion that feels exhausting. The promise of AI was convenience, creativity, and efficiency. But the shadow side is a world where truth has become a negotiation, and where our own minds are the battlefield.
2. The Machinery of Deception
To understand how AI is fanning the flames of misinformation, you have to understand how easily the machinery works. In the past, creating a convincing lie required at least some skill and effort. You needed to know how to write, how to edit photos, how to make a video that would not fall apart under scrutiny. Now, a simple text prompt can generate an entire news article, a deepfake video of a public figure, or a synthetic voice that sounds exactly like your friend or your mother calling for help. CityNews Montreal’s examination of this topic highlights a troubling truth: the cost of creating misinformation has dropped to almost zero, while the cost of verifying truth has risen dramatically. AI models have been trained on vast amounts of human language, images, and sounds, so they mimic not just facts but also tone, sarcasm, cultural references, and emotional heat. They know that people are more likely to share content that makes them feel anxious, righteous, or protective of their in-group. So they generate headlines that exploit those emotions. They know that we skim, that we rarely read past the first few lines, so they put the lie in the title and bury the correction in the body. They know that we trust faces, especially familiar ones, so they place a deepfake of a trusted news anchor in front of a fabricated report. The result is an online ecosystem where misinformation spreads faster than true information, partly because it is designed to be more sensational, and partly because the algorithms that govern our social feeds are optimized for engagement, not accuracy. The report reminds us that these platforms are not neutral; they reward outrage, novelty, and shock. AI simply gives those platforms an endless supply of convincing raw material. For every lie that gets fact-checked, a hundred more are born. For every deepfake that is debunked, another one goes viral before the truth has a chance to catch up. It feels like fighting a hydra, but in this case, the hydra is made of code, and it never sleeps.
3. Real-World Harm, Not Just Digital Noise
It would be easy to dismiss all of this as online noise, but the consequences are profoundly real. In the CityNews Montreal report, the conversation moves beyond abstract concern to tangible harm. False stories about health treatments have sent people to the hospital or made them refuse lifesaving vaccines. Deepfakes have been used to ruin reputations, to extort money, and to create nonconsensual pornography that destroys lives. Fake audio of family members has been used in emergency scams, where an elderly person receives a call that sounds like their grandchild, pleading for bail money or a wire transfer. And in the political realm, AI-generated content has already been used to suppress voters, spread panic, and amplify false narratives about elections. Montreal has not been immune. Local communities, whether they are immigrant groups, francophones, anglophones, or marginalized populations, have seen targeted misinformation that stirs suspicion and division. A fake image of a burning building in a Montreal neighborhood can trigger fear and hate before anyone realizes it was generated by an AI that had no connection to reality. The harm is not just to the individual who is deceived, but to the social fabric itself. When lies flourish, neighbors stop trusting neighbors. When we cannot agree on a basic set of facts, democratic debate becomes impossible. The report paints a picture of a world where every family table conversation can be poisoned by a piece of AI-generated content that someone saw online and passed on with good intentions. Parents share warnings about non-existent dangers. Activists spread manipulated videos of their opponents. Well-meaning citizens amplify scams because the alert seems useful. And once a lie has entered the mind, it is incredibly difficult to remove, even after it has been fact-checked and debunked. This phenomenon, sometimes called the “continued influence effect,” means that reminders of a falsehood can actually make it stronger. The human cost is measured in broken relationships, wasted money, lost trust in institutions, and a creeping loneliness that comes from living in a world where nothing feels solid.
4. The Labors of Truth-Tellers and the Limits of Machine Detection
In response to this crisis, a new industry of truth-tellers has emerged. Fact-checking organizations, activist groups, and responsible journalists are working around the clock to expose falsehoods. Technology companies, meanwhile, are developing detection tools that aim to watermark AI-generated content, identify synthetic voices, and flag deepfake videos. But as the CityNews Montreal piece explains, the race is not promising. Every detector can be fooled by a more sophisticated generator. Every watermark can be stripped by a simple editing program. And even when content is accurately identified as AI-generated, the platforms often fail to remove it because of free speech concerns, legal liabilities, or simply because moderating millions of posts is an unwinnable battle. More troubling is the psychological limitation of detection. If a deepfake is clearly labeled, people may still believe it, especially if it aligns with what they already think. In some cases, labeling misinformation gives it a kind of legitimizing aura, making it seem important and worth sharing. The report suggests that we are entering an era of “information ambiguity,” where the origin of a message is no longer a reliable guide to its truth. Even when a trusted journalist or expert certifies a piece of content as real, there is a lingering doubt. The tools of verification are no longer accessible only to the elite; but the skills to use them effectively are not evenly distributed. Some people are tech-savvy enough to reverse-image-search a photo or analyze the metadata of a file. Many more are simply overwhelmed and resign themselves to believing nothing or everything. The burden falls heavily on teachers, librarians, youth workers, and journalists, who must not only provide facts but also teach people how to interrogate their own emotional reactions. In Montreal, where bilingualism and multiculturalism already complicate the media landscape, the challenge is even greater. A rumor in French can evolve differently than a rumor in English, and both can feed on cultural tensions that are centuries old. The fight against AI misinformation is therefore not just a technological problem; it is a deeply human one, requiring patience, empathy, and an honest recognition of our own fragility.
5. The Human Response: Media Literacy, Community, and Care
The most hopeful part of the CityNews Montreal report is that ordinary people are not passive victims. Around the world, and right here in our neighborhoods, a new form of resistance is taking shape. It is not based on complicated algorithms or government regulations alone; it is based on human habits and values. Media literacy is becoming a subject that matters as much as reading and writing. Schools are teaching children to ask simple but essential questions: Who made this? Why are they sharing it? How does it make me feel? What evidence supports it? These questions are not foolproof, but they slow down the automatic sharing that spreads misinformation. Families are creating rules: do not forward anything before checking the source, do not trust calls from unknown numbers, and do not respond with anger to content that is designed to provoke. Community groups are organizing workshops where people practice detecting deepfakes and identifying bots in their social feeds. These efforts do not require technical genius. They require intentionality, conversation, and a willingness to be corrected. The report also emphasizes that emotional intelligence is a crucial defense against AI manipulation. Misinformation works because it taps into our desire to protect, to be right, and to belong. When we recognize those urges in ourselves, we can pause. We can ask a friend, “Have you seen this? It seems strange.” We can double-check a suspicious image instead of liking it instantly. We can choose to consume news from sources that have a proven record of accuracy, even if those sources are less entertaining than the wild claims that fill our feeds. This is not a return to naive trust. It is a wiser, more caring approach to information: one that values truth as a form of love. When we stop sharing falsehoods, we are not just saving our own reputation; we are protecting our neighbors from confusion, fear, and harm. The human response must also include pressing for stronger accountability from technology companies and governments. We can write to elected officials, support regulation that requires transparency in AI-generated content, and demand that corporations prioritize human wellbeing over engagement metrics. None of this is easy, but it is possible.
6. A Shared Future Built on Honesty and Connection
The AI misinformation crisis is not going away. As the CityNews Montreal piece makes clear, this is a permanent feature of our new reality, not a temporary glitch. The same technology that can draft a poem or help a doctor analyze a scan can also generate a thousand fake stories that pollute the public mind. But the conclusion is not despair. The report reminds us that humans have faced enormous challenges before—mass propaganda, deep political divisions, and rapidly changing technologies—and we have always found ways to adapt. The adaptation will require humility, because none of us are immune to deception. It will require courage, because speaking the truth is often unpopular, especially when a lie is more exciting. And it will require community, because no person can verify everything alone. We need journalists who are willing to say “we don’t know yet” instead of filling the void with rumor. We need algorithm engineers who are willing to sacrifice engagement for accuracy. We need politicians who are willing to protect citizens from synthetic fraud even when those laws limit their own ability to use deepfakes for gain. But above all, we need ordinary people who are willing to think of information as a shared resource, like water or air, to be kept clean and accessible for everyone. On a practical level, this means choosing to slow down. Before sharing something, take a breath. Before getting angry at a headline, read the full story. Before trusting a voice on the phone, verify the source. These small acts seem insignificant, but collectively they create a culture of care. In Montreal, a city known for its pride in critical thought, multicultural dialogue, and artistic creativity, there is an opportunity to be a model for the rest of the world. We can prove that it is possible to use AI without being used by it. We can build spaces—online and offline—where honesty is valued more than outrage, where people can disagree without calling each other enemies, and where the truth is not just a list of facts but a way of being together. The flame of misinformation is strong, but it is no match for the steady, patient light of human attention. The question is not whether we can save our information ecosystem. It is whether we will choose to care enough to do it.