Reframing AI ethics in light of Energy Justice

Technical Summary: Reframing AI Ethics in Light of Energy Justice

Problem Statement
The paper identifies a critical gap in the current discourse on Artificial Intelligence (AI) ethics. While the field has evolved through three conceptual "waves"—moving from speculative existential risks to technical algorithmic issues, and finally to sustainability and structural concerns—the existing frameworks remain insufficient. Specifically, the authors argue that the "third wave," which emphasizes the entanglement of AI with infrastructure, governance, and energy systems, lacks a robust normative framework to operationalize this "structural turn."

The problem is exacerbated by the rapid convergence of AI and energy systems. The massive energy and water consumption required by generative AI (GenAI) and data centers is forcing a re-evaluation of energy strategies, including the resurgence of nuclear power and the "Watt-Bit Collaboration" (locating data centers near power plants). The paper utilizes a future-dated narrative (set in 2025–2026) to illustrate how this convergence creates compound risks, such as the asymmetric distribution of environmental burdens and the vulnerability of critical infrastructure during geopolitical conflicts. In this hypothetical scenario, data centers and nuclear facilities are targeted in conflicts such as a war in Iran and the ongoing war in Ukraine. Current AI ethics frameworks, primarily focused on "ethics" (individual conduct and technical fixes), fail to adequately address these systemic, distributive, and geopolitical challenges.

Methodology
This research employs a theoretical and conceptual analysis rather than empirical experimentation. The methodology involves:

1. Literature Review and Synthesis: The authors critically review the "three waves" of AI ethics as proposed by Bolte and Wynsberghe, analyzing the evolution of ethical concerns from the first wave (speculative/AGI) to the second wave (bias/hallucinations) and the third wave (sustainability/infrastructure).
2. Gap Identification: The paper identifies specific areas excluded from the three-wave framework, most notably the military application of AI and the specific ethical implications of AI in warfare.
3. Interdisciplinary Integration: The authors integrate scholarship from "Energy Justice," a field that frames energy concerns through the lens of justice rather than ethics. They examine the "triumvirate of tenets" of energy justice (Distributive, Procedural, and Recognition Justice) and extensions like Restorative and Cosmopolitan Justice.
4. Conceptual Reframing: The core methodological step is the systematic mapping of existing AI ethical issues (across all three waves and military applications) onto the frameworks of Energy Justice to demonstrate how a justice-based perspective offers a more comprehensive structural analysis.

Key Contributions
The paper makes three primary contributions to the field of AI governance and ethics:

1. Identification of the "Missing" Military Dimension: The authors argue that the three-wave framework of AI ethics is incomplete because it largely omits the security and ethical implications of AI in warfare. They highlight that AI-driven warfare, autonomous weapons, and the targeting of AI infrastructure (data centers) present profound justice issues that current ethical models do not adequately capture.
2. Proposal of a Justice-Based Framework: The paper proposes shifting the normative lens of AI from "ethics" (often focused on individual actions and technical solutions) to "justice" (focused on systemic structures, outcomes, and relationships). This shift is motivated by the success of the energy sector in utilizing justice frameworks to address complex socio-technical systems.
3. Operationalization of the "Structural Turn": The authors provide a concrete mechanism for the "structural turn" proposed in the third wave of AI ethics. They map AI issues onto the Energy Justice Triumvirate:
   • Distributive Justice: Applied to the allocation of environmental burdens (e.g., heat, water usage, rare earth mining) and the benefits of AI, as well as the distribution of damages in warfare.
   • Procedural Justice: Applied to governance, transparency in decision-making, and accountability for AI systems (including black-box algorithms and autonomous weapons).
   • Recognition Justice: Applied to the visibility and respect of communities affected by AI infrastructure, the mitigation of algorithmic stereotyping, and the recognition of diverse knowledge systems.
   • Extensions: The paper also notes the potential utility of Restorative Justice (addressing historical harms) and Cosmopolitan Justice (global energy rights) in the AI context.

Results and Findings
The analysis demonstrates that reframing AI ethics through the lens of justice allows for a more holistic understanding of the AI ecosystem:

• First Wave Issues: Concerns regarding AI control and the erosion of human autonomy are reframed as issues of Procedural Justice (accountability and responsibility).
• Second Wave Issues: Algorithmic bias, hallucinations, and deepfakes are reframed as issues of Recognition Justice (stereotyping and epistemic injustice) and Distributive Justice (unequal access to resources and opportunities).
• Third Wave Issues: The environmental impact of data centers, energy consumption, and infrastructure siting are directly mapped to Distributive Justice (burden sharing) and Recognition Justice (invisibility of local communities).
• Military Applications: The use of AI in war is identified as a critical area requiring Distributive Justice (who bears the cost of war), Procedural Justice (meaningful human control), and Recognition Justice (respect for human life and agency).

The paper concludes that while the three-wave model is useful for categorizing the evolution of discourse, it lacks the normative power to address the systemic nature of modern AI challenges. A justice framework, particularly one borrowed from energy studies, provides the necessary vocabulary to address the "constellation" of technologies, infrastructures, and governance that define contemporary AI.

Significance
The authors claim that their significance lies in offering a "better, richer" set of questions and solutions for the AI field. By moving from an ethics-focused approach (which often defaults to techno-solutionism or individual responsibility) to a justice-focused approach, the field can better address the structural and systemic inequalities inherent in AI deployment.

The paper argues that as AI becomes increasingly embedded in global energy systems and geopolitical strategies, the distinction between AI and energy discussions is no longer tenable. Therefore, adopting the "triumvirate of tenets" from energy justice is not merely an academic exercise but a necessary step to meaningfully operationalize the structural turn in AI ethics. This reframing is presented as essential for addressing the compound risks of AI in the context of climate change, energy security, and geopolitical conflict, utilizing future scenarios to illustrate these potential trajectories.