A recent contribution to The Guardian’s First Dog on the Moon column suggests that facial recognition technology should be deployed widely to combat a perceived wave of shoplifting in supermarkets.
What the Proposal Says
The column argues that supermarkets like Coles and Woolworths could use facial recognition to identify shoplifters in real time.
It claims this technology would ‘fix’ the issue of shoplifting, suggesting it is a practical and scalable solution.
Supermarkets’ Promises and Limitations
Coles and Woolworths have stated they will not abuse the technology if deployed.
However, the column notes that facial recognition algorithms have known biases—particularly against people of colour, women, and certain ethnic groups.
These biases may lead to misidentifications, disproportionately affecting marginalized communities.
How Facial Recognition Works in Retail
- Cameras capture images of individuals entering or moving through store spaces.
- Software compares faces to databases of known individuals, often linked to purchase records or previous incidents.
- Algorithms flag individuals based on match likelihood, triggering alerts to staff.
While the technology can detect presence, it does not verify identity with certainty—especially in diverse populations.
Why This Matters
Deploying facial recognition at scale raises serious ethical and legal questions.
It may lead to surveillance overreach, especially in public or semi-public spaces where people expect privacy.
Moreover, if misused, it could reinforce existing inequalities in policing and retail enforcement.
Key Limitations and Open Questions
- Facial recognition is not foolproof. Accuracy drops significantly with variations in lighting, angles, or facial features.
- Historical data used to train algorithms often lacks diversity, leading to skewed performance.
- There is no clear legal framework in Australia or globally that governs how such technology can be used in retail.
- Public trust in such systems remains low, especially among communities historically targeted by surveillance.
What to Watch Next
Regulators and retailers must address transparency and accountability.
Independent audits of algorithmic performance and bias testing should be required before deployment.
Public consultations could help ensure that solutions do not disproportionately impact vulnerable groups.
The column does not offer a solution to the root causes of shoplifting—such as economic hardship or lack of access to goods—but instead focuses on technological fixes.
For readers concerned about privacy and equity, this proposal warrants cautious scrutiny. The technology may help with detection, but it does not solve the underlying issues of retail access and affordability.
More importantly, it raises a broader question: when should technology be used to monitor people in public spaces?
For those interested in the column, readers can subscribe to receive updates on First Dog cartoons via email. Subscribe here.
For merchandise and prints featuring the First Dog on the Moon, visit their official shop.
For deeper exploration of retail ethics and technology, see the original article at The Guardian.
Sources & further reading
AI-generated illustration.
