WA Police Facial Recognition 'Success' Obscures Structural Bias, Absent Indigenous Consent, and Function Creep Precedents
Original framing: “WA Police’s facial recognition trial delivered strong results. But was it an ethical success?” — The Conversation - Global
The trickster Hermes — patron of thresholds and thieves — grins at 'ethical success': the metric that declares victory IS the trap. Erasmus's 'Praise of Folly' exposes how solemn ethical frameworks legitimize power by debating 'how' while 'whether' is decided. Eshu at the crossroads reveals the paradox: the system works 'too well' at expanding police authority, not at catching criminals. Anansi would note the vendor (NEC) wrote the success criteria. The article's question — 'was it an ethical success?' — performs the trickster's inversion: it makes ethics a performance after the infrastructure is built, not a veto before deployment.
WA Police's facial recognition trial deploys the classic trickster script: a 'successful' technical pilot that legitimizes permanent biometric infrastructure before communities — especially Aboriginal nations — can exercise veto.
The 92% accuracy claim obscures the base rate fallacy that guarantees mass false positives against over-policed groups, while NEC's vendor-written criteria exclude Indigenous data sovereignty, trans existence, and spiritual ontologies of the face. Historical precedent is unequivocal: from Bertillonage to ANPR to China's Sharp Eyes, every 'successful' trial expanded surveillance scope without reducing harm to marginalized people. The solution is not better ethics review but structural divestment: redirecting FR funds to Aboriginal community-controlled safety, legislating moratoriums with Indigenous veto, and treating biometric capture as the colonial continuation it is — not a technical upgrade.