Would a transphobe-identifying app using public facial recognition be a good idea?
Activism
So I’ve been thinking about developing an application (initially for use with CCTV or smartphones) that can identify transphobic people using a biometric database. The goal would be to enable individuals and trans-friendly businesses to know when someone with a known record of anti-trans activism or sentiment is nearby — for example, high-profile figures like JK Rowling or judges involved in hostile court rulings, based on data that is in the public domain or under Creative Commons licences (to avoid GDPR and copyright issues as much as possible).
The idea is that, with this tool, people could choose to refuse service or take precautions if someone known to be hostile to trans rights is identified.
I'm planning to:
Use a facial recognition model like YOLO (You Only Look Once), ideally one that can run on a Raspberry Pi , phone or similar low-power device.
Make the entire source code open source.
Keep the database of known transphobes public and open source.
I’m aware this could raise ethical, legal, or even safety concerns — especially around surveillance, accuracy, and potential misuse — so I’d like to hear from the community:
On a scale from 1 to 10, how bad of an idea is this? (1 = good idea with potential, 10 = really bad/irresponsible idea)
I’d really value honest opinions before I go further. Is it worth exploring, or does it cross too many lines?