When most people are asked, “What makes you human?,” the common response is empathy, love, and compassion. While these seem to be the qualities that separate us from the algorithms we build, artificial intelligence, once a reflection of human ingenuity, is now reshaping what those words mean and how humans are answering those questions.
In today’s day and age, talking to AI chatbots like ChatGPT or Gemini is extremely normalized. In fact, in 2025, Common Sense Media found that about 72% of American teens had used AI companions at least once, and over half said they used them regularly. Many described these chats as emotionally satisfying or even comforting. But when a third of teens say talking to AI feels as good or better than talking to a friend, something fundamental is shifting. Machines can imitate listening and understanding, but they can’t actually feel empathy. What’s alarming about these statistics is that we aren’t just experimenting with AI companions. Instead, we’re letting them fill the emotional gaps that used to unite us.
And this erosion of connection has collectively changed us. In politics and the media, AI has supercharged our confusion about what’s real. Millions even believed that Jake Paul’s makeup video was real, until its Sora AI creation surfaced. If it’s that easy to trick millions of people by a simple SoraAI video, what happens when the subject is a politician, a protest, or a war crime? The National Bureau of Economic Research recently found that when people realize how difficult it is to distinguish real from AI-generated images, their overall trust in the news plummets.
That loss of trust feeds into a cycle of doubt, one that AI is perfectly equipped to exploit. As the Washington Post documented, the number of websites spreading AI-generated false stories has grown by more than 1,000% in a single year. And it’s not limited to niche AI fruits eating themselves, crystal paint ASMR, or even Donald Trump’s video of Hakeem Jeffries wearing a sombrero with mariachi music. Every fake article, fabricated photo, AI voice, and deepfake chips away at our idea of a shared reality.
Additionally, this confusion collides with the way that young people already consume information. Business Insider shares how, according to Google’s own data, almost half of Gen Z prefers to use TikTok or Instagram instead of Google to find news, reviews, and answers. These platforms run on algorithms built to maximize engagement, not report accuracy. The result is a hyper-personalized echo chamber that only exacerbates extremist philosophies or reassures a reflection of one’s own beliefs. It’s misinformation that’s tailored to your personality. As Edward Tufte from the Social Network puts it, “There are only two industries that call their customers ‘users’: illegal drugs and software.”
But a democracy can’t function when its citizens no longer agree on basic facts, and AI is only accelerating that breakdown. The more realistic its outputs become, the easier it is to weaponize them. Deepfake videos of political figures, fake campaign messages, and AI-generated ads are all being simplified to easily accessible tools used to manipulate public opinion.
And yet, awareness doesn’t seem to have an effect on behavior. Harvard’s Misinformation Review found that 83% of Americans are concerned about AI spreading false information in elections, but most of us still scroll through algorithmic feeds without questioning what we see. We live in an age where skepticism is constant but judgement is rare.
Although we’re already seeing the effects, there are solutions: teaching digital literacy early, demanding transparency in how AI content is labeled, and rebuilding spaces where human connection drives conversation, not algorithmic prediction. Although we can pass laws that require disclosure of AI-generated media, above all, cultural habits are necessary to prioritize authenticity over convenience.
If we lose the ability to discern truth, we lose that shared emotional reality that binds us together. When we stop reaching out to one another for comfort or friendship, or we stop reaching across political divides for truth, the line between emotional simulation and political manipulation becomes dangerously thin.
What makes us human has never been perfection, but it has and always will be connection; it’s the common understanding of one another despite our flaws. If we hand that over to machines, we risk becoming the very things we built them to imitate: predictable, optimized, and emotionally vacant. The danger of SoraAI, ChatGPT, and the entire digital ecosystem isn’t rooted in misinformation or even the rise of extremist ideologies in shifting political landscapes, but rather its the underlying replacement of human nature itself.
