Why Gen Z’s ChatGPT Boycott Backfires
Virtue signaling over AI is sabotaging students academically and professionally
A recent survey of Cambridge undergraduates found that 70% have never used ChatGPT for assessed work (or so they claim) and almost 40% haven’t used it for university at all. Their stated reasons? Environmental damage, worries about academic integrity, distrust of Big Tech, and quality concerns. In their minds, boycotting generative AI is the righteous thing to do—proof they care about the planet and the purity of scholarship.
Scratch the surface and those positions collapse. The numbers show Gen Z is aiming at the wrong target and in the process setting up a world where they—not the technology—get shut out. If the boycott sticks, this cohort will graduate with weaker study habits and a skills gap just as employers make AI fluency mandatory.
So let’s address their concerns one-by-one:
Environmental Myths Debunked
Climate guilt over ChatGPT rests on outdated 2022 stats. A prompt now sips ~3 Wh (0.3 Wh on the newest servers) and evaporates only a few milliliters of water. Five prompts use about the energy of one minute of Netflix. A 2024 Nature life‑cycle analysis flips the script: writing a 250‑word page yourself releases ~1.4 kg CO₂e, while ChatGPT produces just 1‑11 g—a 100‑ to 1 500‑fold cut. Boycotting AI therefore increases emissions; the real solution is cleaning up the grid, not banning prompts.
Academic Benefits Proven
Avoiding ChatGPT because you’re concerned it makes you dumber is counterproductive. A World Bank pilot in Nigeria, for example, showed six weeks of chatbot tutoring delivered nearly two school‑years worth of learning gains, with girls gaining more than boys. AI, used well, is an educational accelerant, not a crutch.
Anti‑Big Tech Sentiment Mistargeted
OpenAI, Anthropic and the open‑source community are directly competing with entrenched Big Tech companies like Google. Shunning ChatGPT while defaulting to Google search simply reinforces Big Tech dynamics students claim to dislike. And if corporate control bothers them, they can run open‑source AI models locally.
Quality Concerns a Skill Issue
State‑of‑the‑art models hallucinate < 1 % when summarizing facts, and tools such as ChatGPT’s Deep Research return citations so you can fact check. Mature AI users know which tools to use and how to ensure their accuracy, which is a skill you need to build, like checking academic sources. Choosing to not build that skill, while blaming the technology, is a copout.
The Real Risk: The Coming Hiring Filter
While students focus on unfounded concerns about ChatGPT, they’re missing a bigger real risk: Becoming unemployable. Shopify’s April 2025 memo spells it out: “AI usage is now a baseline expectation” and headcount requests must prove a human can outperform AI before hiring. Recruiters are already scanning resumes for generative AI experience. Entry‑level roles will feel that change first.
Teachers Fail, Students Suffer
Gen Z’s ChatGPT boycott aims for moral high ground yet lands on the wrong side of the numbers, the science, and soon, the job market. If they stay the course, they’ll graduate higher‑emitting, lower‑skilled and less employable than peers who embraced AI. That’s not principled activism; it’s self‑sabotage.
This said, we can’t just blame the students. Yes, they bear responsibility for fact‑checking their beliefs. But universities (and lower grade educational institutions) share the blame for failing to integrate modern tools into pedagogy and for not teaching how to use AI ethically and effectively. The result is a vacuum filled by fear, hearsay, and performative eco‑posturing.
AI isn’t the enemy here, misinformation is. Let’s fight that instead.
AI usage notes: I wrote this with extensive help from ChatGPT, specifically the o3 model. I used it to summarize key points in articles that I had previously read. Then, I gave it my perspective, and an outline of the article I wanted to write, and had it organize the information and create a first draft. I then read through it top to bottom, refined it, and added references. Also, I used ChatGPT’s image generator to create the image at the top of the post.