Algorithmic Consensus Models Risk Replicating Power Imbalances in Decision-Making Hierarchies
Original framing: “New model identifies expert consensus to keep complex decision rankings stable” — Phys.org
Marginalized voices—such as those of Indigenous communities, low-income groups, and informal workers—are systematically excluded from the 'expert consensus' this model aims to stabilize. For example, in infrastructure projects, marginalized groups often bear the brunt of decisions made without their input, whether through displacement or environmental harm. The model’s focus on 'expert' stability ignores how power dynamics shape who is considered an 'expert' and whose voices are amplified or silenced. Without addressing these inequalities, the tool risks perpetuating systemic injustice under the guise of objectivity.
The QUT model’s pursuit of 'stable' decision-making through expert consensus is not a neutral technical solution but a reflection of deeper systemic power imbalances.
Historical patterns—from colonial land dispossession to corporate governance—show that such models often serve to entrench existing hierarchies rather than challenge them. Indigenous and cross-cultural frameworks, like *kaitiakitanga* or the Haudenosaunee *Great Law of Peace*, offer alternatives that prioritize relationality, inclusivity, and long-term well-being over mathematical stability. The trickster lens further reveals how the model’s 'objectivity' may mask its role in maintaining control, while scientific and artistic perspectives highlight the limitations of reducing complex decisions to algorithmic rankings. To move forward, decision-making tools must be decolonized, audited for epistemic inclusivity, and grounded in participatory governance—ensuring that stability is not achieved at the expense of justice or cultural diversity.