0
0
By El Bancroft
[Originally published in the February Issue, 18th of February 2026]
On December 16, 2025, I published a zine titled Why You Need to be More Anti-Gen AI, and some of the information I provided there, I will include here.
Generative AI has been an issue that I’ve been passionate about for a while now, and it still shocks me how normalised it’s become in our society. It is used in our places of employment, in our artistic and creative industries, in our schools and by our governments, often without any thought about its negative impacts.
Gen AI needs data to run, and data needs to be stored in a physical place. These physical places are called Data Centres. According to the Environmental and Energy Study Institute, large data centres can consume up to 5 million gallons per day, and each 100-word AI prompt is estimated to use approximately one bottle of water. Not only that, they also poison water sources for local communities, with residents reporting water being cloudy, brown and unsafe to drink.
These data centres use a massive amount of energy, with power usage projected to rise to around 2967 trillion watts an hour by 2030 (Bush, 2025). In a video by More Perfect Union, in certain areas, residents’ electricity bills have risen to pay for the power of these data centres in the form of supply charges, most likely due to closed-door deals between AI companies and utility providers.
With 4267 data centres in the U.S., and 284 in Australia, these centres are known to target neighbourhoods of marginalised communities and the working class. One example is Boxtown, a historically Black community located in South Memphis, Tennessee. This is where the supercomputers called Colossus, used to train Grok, Twitter’s AI bot, are located.
Thermal imaging revealed that 33 gas turbines operate on the site, emitting approximately 1200 to 2000 tons of nitrogen oxides annually, which can cause respiratory problems and contribute to smog (Eledroos, 2025). In Boxtown, cancer rates are already four times the national average, and that was before the massive generators were built (Eledroos, 2025). This is an example of environmental racism. Environmental racism often comes from landfills, incinerators and hazardous waste disposals, alongside placing ecological burdens such as mining, oil extraction and other industrial agriculture upon indigenous communities and communities of predominantly people of colour.
Not only that, but Gen AI companies like OpenAI employ slave labour to train and filter their AI tools. In a 2023 article by Billy Perrigo, OpenAI paid Kenyan labourers less than $2 an hour to make ChatGPT less toxic, by feeding the AI with “labelled examples of violence, hate speech, and sexual abuse, and that tool could learn to detect those forms of toxicity” and “filter it out before it ever reached the user. It could also help scrub toxic text from the training datasets of future AI models.”
OpenAI sent snippets of graphic text of things ranging from child sexual abuse to bestiality and murder/torture to an outsourcing firm in Kenya. Many workers spoke about being mentally scarred from this and receiving very little in terms of support from their employer, a company called Sama.
Gen AI can’t create something out of thin air. It has no imagination, no life experiences, and no concept of inspiration. Therefore, for gen AI to make a prompt, it must be trained on datasets (texts, songs, movies, videos, books, etc), often without the explicit consent of the creators of those works, and without any credit or compensation. That thing your AI chatbot generated isn’t a piece of original work; it is a “Frankenstein-like product” of stolen creative works (Whiting, 2025).
Another issue that is beginning to be talked about more frequently is Gen AI, more specifically Grok, being used to commit sex-based crimes. On Twitter, users have been using Grok to create sexually explicit images of women and minors without their consent. Gen AI is also used to create porn videos, with some deep-faking real women and children, once again without any consent.
After learning all this, my question to you is, was the shitty email written by ChatGPT worth it? How about that fugly piece of AI-generated “art”? Or that mediocre piece of “writing”? This may sound harsh, but I truly have no respect for people who use Generative AI, not when there’s this much evidence against it.
Generative AI is not a tool for the working class. It isn’t a tool to “make art more accessible for disabled people” (I say this as a disabled person). It is a tool of the fascist elite used to keep us from developing skills on our own, from thinking for ourselves. Fascist governments and billionaires are using Gen AI to spread propaganda (just look at the U.S. Government and ICE). They are using it to create their own narratives.
If you truly care about the environment and your fellow humans, then this is something you must resist.
Sources:
Yañez-Barnuevo, Miguel. Data Centers and Water Consumption. Environmental and Energy Study Institute. June 25, 2025.
Bush, Molly Elise. Construction and Consequences: The Human Impacts of Artificial Intelligence Data Centers. University of Alabama at Birmingham. October 2, 2025.
Eledroos, Nasser. Progress Shouldn’t Poison Black Communities. Tech Policy Press. June 26, 2025.
More Perfect Union. We Found the Hidden Cost of Data Centers. It’s in Your Electric bill. YouTube. August 29, 2025.
Perrigo, Billy. Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic. TIME. January 18, 2023.
Whiting, Sam. AI is generating a ‘Frankenstein-like’ product on stolen creative work. RMIT University. October 9, 2025.
Data Center Map. Australia Data Centers. No Date.
Data Center Map. USA Data Centers. No Date.
Drew Gooden. Greed is Destroying the World. YouTube. December 2, 2025.
Author
Views: 7