Artificial intelligence is increasingly embedded in environmental decision-making, shaping how risks are defined, priorities are set, and regulatory actions are carried out. Current governance approaches tend to emphasize predictive performance, transparency, and ethical principles. Less attention is paid to how algorithmic outputs shape administrative decision-making processes. This policy review assesses artificial intelligence as part of institutional decision-making rather than as neutral analytical infrastructure. Drawing on international governance frameworks published between 2020 and 2025, the analysis assesses how existing policies engage with lifecycle environmental impacts, accountability, and procedural legitimacy. The findings show that risk-based governance and transparency requirements are expanding, yet lifecycle impacts, such as energy use, emissions, and infrastructure dependencies, are not consistently integrated into decision justification and oversight. Accountability mechanisms can show similar limitations. They frequently remain procedural and do not entirely capture how decisions are shaped in practice. Building on the concept of algorithmic sustainability, the review highlights the need to align environmental impacts with institutional responsibility in decision-making. Effective policy design depends on how these relationships are structured, where artificial intelligence influences public outcomes.
Ali et al. (Mon,) studied this question.