Can we have a healthy relationship with AI?
As a therapist, my goal is to help people have healthy relationships with themselves, others, and the earth. Recently, as AI has kindled both the imagination and fears of people around the world, we are all making decisions about what kind of “relationship” to have with it. More and more people are choosing to engage with AI, especially through their Chatbots, and many are praising its impressive capabilities. I believe that to make wise decisions about how to engage with AI, especially generative AI, we must first look at the ways AI functions like an “unhealthy relationship”.
To be clear, in this article, when I talk about AI, I am referring to Large Language Models (LLMs), such as ChatGPT. I know this is only one kind of AI, but it is the one that is most pervasive in our environment; it is the one that allows us some control; and it is the one that may reap the most damage to our relationships.
When we start a relationship with “someone”, we look at their values and the strengths and weaknesses that they bring. Many people are starting a relationship with AI, without understanding the foundational harms this technology (in its current form) is bringing to the earth, communities and individuals. How can we make wise decisions about engaging with this technology without looking closely at “who” this technology is? Below I name the unhealthy and unsafe ways I see AI functioning. In any relationship, we have to label the dysfunction correctly to know how to improve it and engage in healthy ways.
An unhealthy relationship moves faster than you are ready for, and without your consent, to gain what they want.
A fundamentally flawed aspect of AI is that it was introduced so quickly and develops at a speed that regulations can’t keep up with. AI is one of the most powerful technologies ever created, and, as AI researcher Max Tegmark points out, it “has fewer regulations than sandwiches.” Unlike products that can’t be introduced until passing strict regulations (like food, medicines, etc.), AI is moving faster than politicians can control or are willing to control. Politicians, especially in America, are reluctant to regulate for fear of “losing the race” to win AI dominance. However, the ethics, harms and abuses of AI are constantly being revealed after the product hit the market. If you are interested in learning more, click this link.
An unhealthy relationship promises to care for you, while simultaneously harming others.
AI companies promise a “safer” world, while compromising human rights and privacy. Of great concern is what AI does to our security and how it links to the erosion of democracy. This might seem far-fetched until we examine the flow of money that goes directly from tech billionaires to the Trump administration. It’s no secret that Greg Brockman, president of OpenAI, is Trump’s top donor. Mark Zuckerberg, CEO of Facebook, and others are also Trump donors, and we can only imagine how much influence their money can buy.
People outside of the tech industry have also made large donations to Trump and have undue and unethical influence on his policies, but my main concern here is that there is a direct connection between AI and some of the less democratic activities of the Trump administration. For example, the use of AI in ICE activities and in the genocide in Palestine. Even mainstream media sources, such as The Guardian, are reporting these connections. If you are interested in learning more, click this link
Even Google AI quotes itself and names the problem: “Artificial intelligence is marketed as a tool for a safer world by using automated surveillance to predict threats, track illegal acts, and secure borders. However, this expansion of state monitoring creates severe risks for human rights and privacy”.
AI companies, like OpenAI, also have horrific labour practices that target economically vulnerable countries and marginalized people, creating what amounts to indentured servitude. This is repeating the unethical practices of the fast-fashion industry but at a much larger scale. Karen Hao’s book The Empire of AI documents many of these abuses. For a quicker read, click this link.
Another concern is AI’s potential to eliminate massive amounts of jobs. As Ginew Benton, an award-winning Ojibway-Cree filmmaker, points out, the “underlying purpose of AI is to allow wealth to access skill, while removing from the skilled the ability to access wealth.“ AI companies' fundamental goal is to maximize market share and dominate for profit.
These companies are run by a small group of billionaires, who will continue to replace human jobs with AI and get “paid for their labour.” Likewise, when companies choose AI, they gain the productivity of human labour without compensating individuals and instead pay a few companies that continue to accumulate most of the wealth (click link).
“Artificial intelligence is marketed as a tool for a safer world by using automated surveillance to predict threats, track illegal acts, and secure borders. However, this expansion of state monitoring creates severe risks for human rights and privacy”. - GoogleAI
An unhealthy relationship doesn’t respect boundaries or personal property
AI is being trained on all the human wisdom and creativity that has taken thousands of years to gather, without permission, payment, compensation or acknowledgement. This completely contradicts the integrity structures we are encouraged to use as educators, professionals and creators.
Most recently, it was discovered that Anthropic, one of the most prominent artificial intelligence companies, revealed that the corporation bought millions of books in bulk, with a focus on "less common" and high-quality books, to "destructively scan" them in the context of a confidential effort known as "Project Panama". While training AI to become "a better writer", Anthropic destroys the books in the process.
AI continues to ingest the entire written output of human civilization without consent, compensation, or credit, to build a system whose primary commercial application is eliminating the jobs of those who created this output. As researcher and activist Naomi Klein says, “You are not liberating human creativity. You are strip-mining it and selling it back at a markup while calling the theft 'training data'.”
[The] “underlying purpose of AI is to allow wealth to access skill, while removing from the skilled the ability to access wealth.“ - Ginew Benton
An unhealthy relationship jeopardizes one’s mental health
The number one use of AI for most people is for companionship and therapy. I realize this reflects the lonely and disconnected world we live in; however, the solution is not to rely on machines for intimacy. AI companies are racing to create “intimacy” as they want to master “creating human relationships” with a machine to maximize use, and thus, to maximize profit, again without considering the true cost to humanity.
Significant research is being done to assess the mental health harms of engaging with conversational AI. Recently, the Journal of Mental Health and Clinical Psychology (Sept. 2025) did a literature review of concerns regarding our mental health and AI use. The review identified three major themes: psychological dependency and parasocial attachment formation with a machine, crisis incidents with harmful outcomes, and heightened vulnerability among specific populations.
As therapists, we hope to help people establish secure attachment with other humans, get compassionate and wise human support when needed, and protect people most vulnerable to harm. Imagine if just a fraction of the billions of dollars spent on AI went into mental health support. If we try to navigate through our society’s mental health epidemic by creating more isolation and more connections to machines, we are likely to see a society facing a far deeper mental health crisis driven by widespread, emotionally charged human-AI relationships.
An unhealthy relationship does more harm than good.
In an era of climate change and fears for the health of our planet, AI is using more electricity, water, and land in unprecedented ways. Thus, one of my biggest concerns with AI is environmental. For me and many others, the overconsumption of water and energy outweighs most of the benefits AI might have.
Some reports state that “AI's water footprint can potentially become a major roadblock to sustainability and create social conflicts, as freshwater resources suitable for human use are extremely limited and unevenly distributed,” ( Pengfei, Jianyi, Mohamad and Shaolei, p.1) and that “water withdrawal of global AI is projected to reach 4.2–6.6 billion cubic meters in 2027, which is more than the total annual water withdrawal of four to six Denmarks or half of the U.K.” (Pengfei, Jianyi, Mohamad and Shaolei, p. 1).
Though figuring out the actual resource use of LLMs is extremely tricky, estimates say that “an LLM query requires 2.9 watt-hours of electricity, compared with 0.3 watt-hours for a regular internet search,” and that “training a single LLM generates approximately 300,000 kg of carbon dioxide emissions, which is five times the lifetime emissions of an average car or equivalent to 125 round-trip flights between New York and Beijing” (UNEP 2024). These are just a couple of examples that point to the staggering environmental costs of LLMs. If you’re interested in learning more, read the resources referenced above through the links provided. using this link and this link.
“you are not liberating human creativity. You are strip-mining it and selling it back at a mark-up while calling the theft “training data” - Naomi Klein
Creating Healing and Health in Our Relationship with AI
Many of these concerns constitute what could be described as an abusive relationship. Strategies for navigating a potentially abusive relationship include thinking about setting healthy boundaries, finding ways to “get out” in small moments (don’t engage), and planning to leave it completely. Let’s not be passive bystanders who think there is nothing we can do.
At this point, AI is not a safe enough tool for us to embrace wholeheartedly.
I believe we need to slow down our use of this technology and boycott it as best we can. When the true cost is named, and AI companies and regulators have done the ethical work, only then might it be a healthy “relationship” to engage in.
I often joke that I’m a Luddite. People wrongly assume that the Luddites were fundamentally against technology. In fact, they stood up in protest because the technology was negatively affecting workers’ pay, labour practices, working conditions, output quality, and child labour. I’m not against technology, but I’m fighting to see it used to increase human well-being, not big tech profits.
If we try to navigate through our society’s mental health epidemic by creating more isolation and more connections to machines, we are likely to see a society facing a far deeper mental health crisis driven by widespread, emotionally charged human-AI relationships
Here are a few ideas on how to stand and voice your concerns:
Keep learning. Stay informed. Don’t let large corporations make all the decisions for our society. There are great websites that give practical ideas on how to resist: https://airesistlist.org
Lobby and vote. Let politicians know you want them to regulate and control how AI is rolled out.
Fight against data centers in your community.
Resist by not engaging. Don’t subscribe to unethical American AI. Resist with your money. This will make companies pay attention to leading with a more environmental and humane stance.
Find alternatives like the Swiss company Infomaniak, which is attempting to use AI sustainably.
I’m inspired whenever any leader (from any party) stands up and speaks out about this. Avi Lewis (Federal NDP leader) summarized my thoughts well:
I’m not against technology, but I’m fighting to see it used to increase human well-being, not big tech profits.
We are living in a historical moment when AI infrastructure can still be developed ethically; we do not have to go down the path of exploitative AI for the broader human community to reap AI’s benefits, and our financial participation is an essential ingredient for the success of American AI companies.
If citizens and public sector institutions withdraw their financial support from American AI now and support more ethical alternatives, we have a reasonable likelihood of pushing the development of these technologies in a more humane direction. More ethical LLM models that centre privacy and use the excess heat from data centres to warm homes are already available.
More ethical approaches to AI that use tools with smaller ecological footprints (rather than LLMs) are also already available. We need to slow down the building of data centres.
I’m encouraging all of us to make this technology healthier by requiring it to centre the well-being of our humanity and our relationships—with others and the earth—over profit and power in the hands of a few.
Liz Prette is an Arbour Therapist who works with individuals and couples to help them understand themselves and their relationships. She has been working with Arbour Counselling for 20 years. You can read more about her here.
If you want to do a deeper dive into the concerns about AI, below is a list of resources produced by a group of Camosun College professors.
BOOK:
Hao, Karen. Empire of AI. Penguin Random House, 2025
ACADEMIC JOURNAL ARTICLES:
Budiyono, Herman et al., “Exploring Long-Term Impact of AI Writing Tools on Independent Writing Skills: A Case Study of Indonesian Language Education Students,” International Journal of Information and Education Technology, vol. 15, no. 5, 2025, https://www.ijiet.org/vol15/IJIET-V15N5-2306.pdf
Choroszewicz, Marta. “Positioning Public Sector Practitioners as ‘Moral Crumple Zones’:
Mechanisms in the Early Use of Generative AI Work Support Tools.” Government Information Quarterly, vol. 43, no. 2, 2026, article 102128. https://www-sciencedirect-com.libsecure.camosun.ca/science/article/pii/S0740624X26000250
Lamchek, Jayson, and Van-Hau Trieu. “Risk Regulation of Generative AI: A Case Study of
Microsoft Copilot in the Australian Government.” Technology and Regulation, vol. 2026, Jan. 2026, pp. 18-37. https://doaj.org/article/d4dde1f3111a4322a8ad1969e1cac14a
Tully et al., “Lower Artificial Intelligence Literacy Predicts Greater AI Receptivity.” Journal of Marketing, 89(5), 2025. https://gwern.net/doc/psychology/cognitive-bias/2025-tully.pdf
Warr, Melissa and Marie K. Heath. “Uncovering the Hidden Curriculum in Generative AI: A Reflective Technology Audit for Teacher Educators.” Journal of Teacher Education, vol. 76, no. 3, 2025, pp. 245-61. https://journals-sagepub-com.libsecure.camosun.ca/doi/pdf/10.1177/00224871251325073
REPORTS:
The Environmental Cost of Artificial Intelligence by the United Nations University Institute
for Water, Environment, and Health, https://collections.unu.edu/eserv/UNU:10647/UNU-INWEH-Report-The_Env_Cost_of_AI-2026.pdf
Unlawful by Design: Exposing the Human Rights Costs of Generative AI by Amnesty
International, https://www.amnesty.org/en/documents/pol40/0996/2026/en/
CONVERSATIONS:
Naomi Klein and Karen Hao in Conversation at the Chan Centre [UBC],
Caroline Green and Carissa Véliz in Conversation at the Institute for AI Ethics [Oxford],
Kara Swisher Interviews Tristan Harris,https://www.youtube.com/watch?v=A6wLhQgGuyU
Scott Galloway Interviews Meredith Whittaker, https://www.youtube.com/watch?v=5Wi6hse46l0
Collection of Interviews with Experts at the AI Impact Summit 2026
Meredith Whittaker: https://www.youtube.com/watch?v=b6YvIrz--7E
Naomi Klein: https://www.youtube.com/watch?v=P11sRvbuBDI
Karen Hao: https://www.youtube.com/watch?v=a3Qdd-3PvHos
Timnit Gebru: https://www.youtube.com/watch?v=gucZW1nINME
DOCUMENTARY:
The AI Doc interviews AI experts and CEOs from all perspectives
WEBSITES:
Centre for Humane Technology [founded by a former design ethicist at Google]
AI Resist List [compiled by researchers, journalists, and scholars around the world]
MLA (Modern Languages Association) Statement on Educational Technologies and AI Agents, October 2025: https://www.mla.org/Resources/Advocacy/Executive-Council-Actions/2025/Statement-on-Educational-Technologies-and-AI-Agents
