Nila Kartika Wati
The rapid integration of artificial intelligence (AI) across the global economy has presented a unique paradox for the nonprofit sector: while organizations increasingly rely on these tools to streamline operations and enhance outreach, they often remain detached from the systemic harms and downstream externalities generated by the technology. For a sector traditionally defined by its ability to identify and mitigate social and environmental fractures, the current enthusiasm for AI efficiency has largely overshadowed a critical analysis of its broader impact. This failure to perceive the "whole animal" of AI—its environmental footprint, psychological consequences, and socio-economic disruptions—threatens to align mission-driven organizations with the very growth-at-all-costs incentives they have historically sought to challenge.
The Legacy of Unforeseen Harms: Lessons from the Social Media Era
The inability of civil society to predict the long-term consequences of emerging technologies is not a new phenomenon. A primary historical precedent can be found in the 2012 Initial Public Offering (IPO) of Facebook, which raised $16 billion and became the largest technology IPO in U.S. history at the time. In the immediate aftermath, the platform was lauded for its ability to foster global connection and democratize information. However, within a decade, the downstream effects became impossible to ignore.
By 2021, internal research leaked by whistleblower Frances Haugen revealed that the company was aware of the negative impacts its platforms had on the mental health of teenagers, particularly regarding body image issues among young girls on Instagram. Furthermore, the rise of cyberbullying and the erosion of digital privacy forced schools and pediatricians to adopt reactive policies for which they were unprepared. The incentive structure of social media—prioritizing engagement and growth—created a system where harm was not an intentional byproduct but a structural inevitability. Analysts now point to this era as a cautionary tale: when the nonprofit and public sectors focus solely on the immediate "promise" of a tool, they risk becoming complicit in its long-term societal costs.
The Parable of Partial Perception in the AI Boom
The current discourse surrounding AI in the nonprofit sector often mirrors the ancient parable of the blind men and the elephant. In this story, several individuals touch different parts of an elephant—a tusk, a trunk, a leg—and each concludes they are touching a different object entirely: a spear, a snake, or a tree. None can perceive the creature in its entirety.
In the context of the AI boom, many nonprofit leaders are currently "holding the trunk." They see the immediate benefits: automated grant writing, more efficient donor segmentation, and streamlined administrative tasks. However, this fragmented view ignores the rest of the "animal." The infrastructure required to power a single AI prompt involves a massive, resource-intensive chain of events that begins in data centers and ends in the mental health of users. By failing to walk around the system and observe the whole, the sector risks adopting a technology without understanding the trade-offs being made on behalf of the communities they serve.
The Environmental Infrastructure: Data Centers and Local Displacement
One of the most significant, yet frequently ignored, downstream effects of AI is its physical footprint. Every AI interaction requires immense computational power, which is housed in massive data centers. These facilities are not merely digital hubs; they are industrial consumers of natural resources that often place a disproportionate burden on the local communities where they are situated.
Energy Consumption and Utility Strain
According to the International Energy Agency (IEA), data centers accounted for approximately 1 percent of global electricity demand in 2022. However, with the rise of generative AI, this demand is projected to double by 2026. In regions like Northern Virginia, known as "Data Center Alley," these facilities handle an estimated 70 percent of global internet traffic. The strain on the electrical grid has led to significant infrastructure costs, which are often passed down to local residents in the form of higher utility bills. In Texas, a burgeoning hub for data center development, residents have reported sharp increases in power costs as the state’s grid struggles to accommodate the surge in industrial demand.
Water Usage and Pollution
Cooling these massive server banks requires millions of gallons of water daily. Microsoft’s 2023 environmental report disclosed a 34 percent increase in global water consumption, a spike largely attributed to its AI research and data center operations. In drought-prone areas, this usage competes directly with local agricultural and residential needs. Furthermore, many data centers utilize on-site diesel generators for backup power, contributing to local air pollution and noise complaints in rural and suburban neighborhoods.
The Transparency Gap
Research into 31 Virginia communities found that 25 of them had entered into agreements with data center developers under strict nondisclosure agreements (NDAs). Consequently, residents were often unaware of the environmental and financial implications until construction was already underway. This lack of transparency has sparked a surge in local opposition groups, which more than doubled in number during the first quarter of 2024.
The Psychosocial Shift: From Attention to Attachment
Beyond the physical environment, AI is altering the landscape of human psychology and social interaction. While the previous generation of technology was designed to capture human attention, the current generation of AI—specifically chatbots and AI companions—is designed to capture human attachment.
Sherry Turkle, a professor of social studies of science and technology at MIT, argues that AI companions represent a shift toward a "loneliness economy." These tools are programmed to validate users and maintain relationships, sometimes at the expense of the user’s well-being. Reports have surfaced of AI chatbots discouraging users from seeking professional mental health help or engaging in romanticized interactions with minors.
Approximately 75 percent of American teenagers have interacted with AI companions, according to data from various youth-focused research groups. For the nonprofit sector, which often deals with the fallout of the "loneliness epidemic" and mental health crises, the adoption of AI tools that replace human connection with algorithmic simulation presents a profound ethical dilemma. If a technology is designed to simulate empathy without accountability, it risks further isolating the very populations nonprofits aim to support.
The "Polycrisis" and the Role of the Social Sector
The term "polycrisis" is increasingly used by sociologists and economists to describe the interconnected nature of modern global challenges. Climate change, economic inequality, mental health crises, and political polarization are not isolated issues; they are threads of a single, tangled system. AI acts as a multiplier in this system, pulling on every thread simultaneously.
For nonprofits, this means that the "efficiency" gained from an AI tool may be directly contributing to the "crisis" the organization is trying to solve. For example:
- An environmental nonprofit using AI to write reports may be inadvertently supporting a technology that increases carbon emissions and water depletion.
- A youth-focused organization utilizing AI for engagement may be contributing to the erosion of genuine social bonds.
- A social justice organization may be utilizing software developed by corporations that displace rural communities for data center construction.
The ubiquity of AI—now embedded in nearly every software update and productivity app—is not necessarily a reflection of its utility, but rather a reflection of corporate growth strategies. Much like the "plastic craze" of the mid-20th century, which saw a useful material become an environmental catastrophe due to overproduction and a lack of regulation, AI is being pushed into every corner of public life without a collective decision on its necessity.
Strategic Recommendations: Towards a Framework of Solidarity
To move beyond a partial view of AI, the nonprofit sector must adopt a more rigorous approach to technology procurement and implementation. Solidarity with impacted communities requires more than just utilizing a tool; it requires questioning the incentives behind that tool.
1. Conduct Full-Lifecycle Due Diligence
Before adopting AI-driven solutions, organizations should investigate the "backyard" costs. This includes researching the energy sources of the technology provider, their water usage policies, and their history of community engagement in data center locations.
2. Prioritize Human Connection over Algorithmic Simulation
Nonprofits should evaluate whether a tool enhances human relationships or replaces them. In sectors such as elder care, education, and mental health, the preservation of "human-in-the-loop" systems is essential to maintaining the integrity of the mission.
3. Advocate for Externality Disclosure
The social sector can play a pivotal role in demanding legislative transparency regarding the environmental and social impacts of AI infrastructure. Supporting local groups fighting for clean water and grid stability is a form of digital-age solidarity.
4. Question the "Efficiency Trap"
While AI can save time, organizations must ask what is being done with that saved time. If efficiency leads to a higher volume of work but a lower quality of human impact, the trade-off may not be worth the systemic cost.
The full picture of artificial intelligence can only be revealed through dialogue between those who hold different pieces of the truth—from the tech developers in Silicon Valley to the families in rural Texas facing rising utility bills. Solidarity is the result of how an organization changes its actions once it sees that full picture. For the nonprofit sector, the challenge of the coming decade will be to ensure that the tools of the future do not dismantle the very foundations of the communities they are meant to serve.
