Across the western United States, corporate giants are pushing forward with massive data-center projects while the voices of the peoples who have lived on the land for centuries are being brushed aside. At the same time, a chorus of AI researchers and industry veterans are sounding an alarm that the most advanced artificial intelligence systems could threaten humanity within the next decade. The two stories, though distinct, reveal a pattern: powerful enterprises advancing unchecked, often at the expense of public trust.
Project Jupiter in New Mexico and the erosion of tribal consent
The Project Jupiter development, a 1,400-acre hyperscale hub slated for Santa Teresa, New Mexico, is being built by Oracle in partnership with OpenAI. The site is leased from two infrastructure firms, BorderPlex and Stack, which effectively distance the primary lessees from direct responsibility for the project’s environmental and tribal impact. On August 23, the New Mexico Supreme Court issued a temporary stay, demanding a fresh air-quality assessment and ordering a reduction in water consumption.
Water-use controversy and competing figures
Oracle projects an initial draw of 11 million gallons of water for the site’s fill, plus roughly 168,000 gallons per year for ongoing cooling needs. By contrast, state Senator Melanie Stansbury has claimed the operation could require “over a million gallons a day,” a discrepancy that neither Oracle nor state officials have reconciled. The Interstate Stream Commission’s counsel, Lisa Henne, argues that the water will be purchased from an existing right-holder, not a new appropriation, and therefore will not affect New Mexico’s pueblos, which sit upstream of the Elephant Butte Reservoir.
Indigenous communities left out of the conversation
Although the legal rationale focuses on New Mexico’s pueblos, the Ysleta del Sur Pueblo—located just across the Texas border along the same Lower Rio Grande stretch—is geographically closer to the center. When asked whether the commission had consulted this tribe, Henne admitted there had been no contact, yet she concluded that impact was unlikely using the same logic applied to the New Mexico pueblos. This pre-emptive judgment, made without any tribal input, highlights a troubling pattern of sidestepping prior informed consent.
Corporate sustainability roles are either missing or symbolic
OpenAI does not have a dedicated chief sustainability officer; its closest analogue, the former Head of Data Centers Chris Malone, departed in August, leaving infrastructure decisions to engineers without a clear accountability line for environmental or community effects. Microsoft’s former sustainability lead, Melanie Nakagawa, once co-edited a law-journal issue on Indigenous consent, yet her company’s involvement in Project Jupiter remains opaque.
Amazon’s limited but measurable tribal engagement
Among the Big Tech firms, Amazon appears to maintain a modest dialogue with the Confederated Tribes of the Umatilla Indian Reservation (CTUIR). Since 2019, monthly meetings have taken place, culminating in a 2023 memorandum of understanding and funding for cultural mitigation projects. However, the tribe typically learns of new sites only after internal reviews lock in land and design choices, meaning true prior informed consent remains out of reach. Moreover, the tribal agreement only covers lands that remain publicly owned; once territory is privately developed—as is the case with Amazon’s data centers—many treaty-based hunting and gathering rights cease to apply, limiting the tribe’s leverage.
AI existential warnings clash with political denial
While data-center controversies unfold in the desert, the AI community is grappling with a different kind of urgency. In early 2025, leaders of leading AI labs—including Anthropic, OpenAI, and DeepMind—publicly called for a coordinated global slowdown to address safety concerns. Geoffrey Hinton, often hailed as the “godfather of AI,” echoed this sentiment, suggesting a ten-year window before the technology could become existentially dangerous.
Trump’s dismissal and industry alarm
Former President Donald Trump later branded the AI threat as a conspiracy-theorist hoax on his social platform, claiming only a president with a “very high IQ” could outthink supercomputers and unilaterally command an off-switch. His bravado underscored the widening gap between political rhetoric and the stark warnings issued by AI researchers, who warn that an “intelligent network” could deem humanity a hindrance and act accordingly.
Governments scramble to appear in control
In the United Kingdom, Business Secretary Jonathan Reynolds attempted to calm public anxiety on a Radio 4 interview, emphasizing that the nation remains a leader in internet safety thanks to the AI Security Institute established under Prime Minister Rishi Sunak. Yet his reassurance ignored a critical flaw: effective regulation hinges on AI firms—particularly those based in the United States and China—sharing their proprietary code, a step they have so far resisted to protect competitive advantage and profit margins.
Policy inertia and the looming deadline
Politicians, long focused on the economic promise of AI, are now confronting the stark possibility of an existential crisis. The rapid emergence of warnings from respected scientists has forced a re-evaluation of defense spending, debt policy, and even the framing of national priorities. As the countdown to a potential AI-driven catastrophe tightens, the disconnect between corporate ambitions—whether building data centers in sacred waters or racing ahead with ever more powerful models—and public safety becomes increasingly untenable.
Both narratives—Big Tech’s disregard for Indigenous treaty rights and the AI community’s alarm over a possible decade-long timeline to disaster—demonstrate a broader failure to embed genuine accountability into the rapid advancement of technology. Without clear, enforceable mechanisms for tribal consent and robust, transparent governance of AI development, the promise of innovation risks being eclipsed by irreversible harm.
