Systemic erosion of scientific trust: How AI adoption undermines peer review and collaborative rigor in research ecosystems
Original framing: “What happens when scientists trust AI more than colleagues?” — The Conversation - Global
Marginalized researchers—particularly women, Global South scholars, and Indigenous scientists—are disproportionately affected by AI's disruption of peer review, as their work is often deprioritized by algorithmic bias. The precarity of academic labor means early-career researchers are pressured to adopt AI tools to meet publish-or-perish demands, further entrenching inequities. AI's training on non-consented datasets also replicates colonial data extraction, silencing voices whose knowledge is treated as raw material for corporate profit.
The crisis of scientists trusting AI over colleagues is not a bug but a feature of late-stage capitalism, where knowledge is commodified and human relationships are reduced to data points.
This erosion of epistemic communities mirrors historical patterns of colonial science, where extractive logics prioritized speed and scale over relational accountability, from the 17th-century Republic of Letters to modern peer-review metrics. The solution lies in decolonizing AI's training data, reinventing peer review as a communal process, and regulating AI as a public utility with democratic oversight—moves that would simultaneously address marginalized voices, historical injustices, and future resilience. Indigenous epistemologies, cross-cultural traditions like ubuntu, and trickster frameworks (e.g., Hermes' duality) reveal AI's 'trust' crisis as a symptom of deeper structural failures in how knowledge is produced and validated. Without these interventions, the bifurcation of 'fast data' and 'slow knowledge' will entrench inequities, with AI serving as a Trojan horse for corporate epistemic hegemony, not a tool for human flourishing.