Denial, Bargaining, Anger: Where Our Relationship With AI Goes Next

Nearly four years after ChatGPT’s public debut, our collective relationship with AI has followed a pattern that looks a lot like the five stages of grief. What we’re mourning is the familiar, along with a sense of stability that was already hanging on by a thread before any chatbot arrived. The stages haven’t moved in a straight line, and traces of each keep resurfacing.

Early on, plenty of us were curious about what the technology could do, while skeptics denied it would ever make a real impact. Then AI moved into our daily lives and workflows, and the bargaining set in. “It will only handle the boring tasks.” “A few weeks without it won’t hurt.” “Maybe I can skip it altogether.” I’ve watched that bargaining play out in my own industry, where agency budgets keep shrinking while client demand stays the same, and AI keeps getting pitched as the way to close the gap. Running underneath every stage has been the same anxiety and fear. Now that fear is boiling over into anger.

Why the Backlash Is Arriving Now

Anger tends to show up when fear has nowhere left to go. Three things have pushed us to that point.

The first is ubiquity. A Pew Research Center survey of more than 5,000 U.S. adults found Americans are far more worried than excited about the growing role of AI in daily life. That worry is no longer abstract, because the technology sits inside our search bars, our inboxes, our classrooms, and our job descriptions.

The second is that the stakes have become personal. Early debates about AI were about capability. The current debate is about rent, tuition, and whether our kids will learn to think without a machine doing the thinking first.

The third is volume. Our feeds, ads, and customer service chats are filling up with machine-made content, and more of us can tell. Recognition breeds resentment quickly.

What the Anger Looks Like

Regulators Are Looking at Young Users

Pressure to regulate AI is growing, especially around how young people use it. In September 2025, the Federal Trade Commission launched an inquiry into AI chatbots acting as companions, asking several major companies how they test for and limit harm to children and teens. When a federal regulator starts asking those questions, public concern has moved well past the dinner table.

Educators Are Watching How Students Think

Teachers and professors describe students who hand their thinking off to AI and then struggle to recall what they submitted. The research is still early. An MIT Media Lab preprint titled Your Brain on ChatGPT found that participants who wrote essays with a language model showed weaker neural engagement and poorer recall of their own work. The study was small and had not been peer-reviewed when it was posted, so it’s best treated as a signal and not a verdict. Still, it gave educators language for something they were already seeing in their classrooms.

Workers Are Afraid for Their Careers

Fear about work is the most widespread and the most personal. Pew found that 52% of U.S. workers were worried about the future impact of AI in the workplace, 32% expected fewer job opportunities for themselves, and only 36% felt hopeful. Creative professionals feel it acutely, since their work is exactly what generative tools imitate best.

The layoff numbers add weight. Challenger, Gray & Christmas tracked 1,206,374 announced job cuts in 2025, up 58% from the year before. Not all of those cuts were about AI, and it would be unfair to say they were. But the perception matters as much as the attribution. Entire teams are being trimmed because one person knows how to pull the right levers, and many of us feel like our heads are on the guillotine.

Audiences Are Pushing Back on AI-Generated Content

That fear for our own futures, and for our children’s futures, is now surfacing as backlash against AI-generated content. When McDonald’s Netherlands released an AI-generated holiday ad, the reaction was fierce enough that, according to the BBC, the company pulled it. Brands that treated synthetic content as a cost-saving shortcut are discovering that audiences read it as a message about how much they’re valued.

The Paradox at the Center

AI has also become unavoidable. Most of us use it every day, and many of us have to just to keep up. A writer can despise what AI is doing to the industry and still run drafts through it at night. A teacher can ban it in class and quietly use it for lesson plans.

A counter-movement is forming around human-made work, basic learning principles, and in-person community, and it’s pushing back against a tool most of its own members still depend on. That isn’t hypocrisy. We can resent something and rely on it at the same time, and that is exactly what grief looks like in practice.

How the Next Year Is Likely to Unfold

No one can forecast this precisely, but I see five developments as likely.

• More rules aimed at young users. Age limits, disclosure requirements, and safety testing for AI products used by minors are likely to expand.
• “Human-made” becomes a selling point. Expect more brands to advertise the people behind their work, and more audiences to ask for proof.
• Schools redraw the lines. More in-person exams, oral defenses, handwritten work, and explicit policies on when AI belongs in an assignment and when it doesn’t.
• Trust becomes a competitive edge. Companies that are open about how AI affects their own employees will earn more goodwill than those that announce cuts and call them efficiency.
• A quieter turn toward slowing down. In-person communities, offline hobbies, and deliberate pauses are likely to grow, not as a rejection of technology but as a way to rebuild focus, including in how we learn.

That last shift matters most in learning, where calm and attention are part of the work and not a separate problem. Schools are already acting on this. Districts across the country have added breathing exercises, short meditations, and yoga to the school day, and the CDC has endorsed mindfulness practices to help students manage stress and emotions. None of this makes mindfulness a cure-all, but it reflects a growing recognition that understanding takes time, and a rushed or anxious mind has a harder time doing it. AI makes that easy to forget. The real risk of AI-assisted learning isn’t that students get answers too easily, but that they skip the slower, more uncomfortable work of understanding why an answer is right. Protecting that work is where companies and educators can start.

What Companies and Educators Should Do

Public trust will be earned through specific choices.

• Protect the team. Be honest with employees about where AI will and won’t be used, and invest in reskilling before announcing reductions.
• Show the sources. Tools that cite where an answer came from help users check the work and keep thinking for themselves.
• Build in a pause. Give learners a moment to settle before a task, and let them see why an answer is wrong, not just that it is.
• Skip the guilt. Notifications designed to shame users into streaks may raise engagement numbers, but they erode the trust that this backlash is now testing.
• Keep humans visible. Label AI-generated content clearly, and credit the people involved in the work.

For educators, the goal isn’t to choose between banning AI and embracing it. It’s to make the purpose of each assignment clear, so students understand what the exercise is meant to build and why a shortcut defeats it.

Keeping the Humanity of the Work at the Center

The anger around AI is easy to dismiss as resistance to progress. I think it’s more accurate to read it as grief, and as a sign that we care about what we stand to lose. Many of us already sense how AI enriches us, personally and professionally, and how it depletes us, professionally, socioeconomically, and even financially. We’re still discovering where that balance lands.

That balance won’t settle on its own. Technology can look neutral, but it’s never without consequence. The way it’s built shapes how it gets used, and some of its effects are unintended, taking years to fully grasp and understand. That’s why it matters whether a tool is designed to take from people or to nurture them. The organizations that earn public trust will be the ones that choose to nurture, and that protect their teams as much as their customers. They’ll keep the humanity of the work at the center, because people are who it’s for.

If we let this grief run its course, what comes after anger won’t be surrender. It will be acceptance built on knowledge: understanding AI well enough to know when it serves us, when it takes from us, and when to set it aside. Turning that knowledge into wisdom is the work of the next year.