Reflection on Facial Recognition on British Passports

AI is rapidly advancing each and every day with many people dreaming of the its great applications in the future of technology. But is AI advancing down the wrong path?

A Tweet by @elainebabey on February 25th highlights just how artificial intelligence may not be the answer to all our problems. Whilst applying for a passport Elaine was notified that her application image was not acceptable because it appeared as if her mouths was open. In actuality the cause of this was due to the artificial intelligence being trained with a data set that Elaine did not fall into based upon her physical features.

Whenever new technology is created one of the first questions that must always be asked is who paid for it and why? This will often reveal much more about the technology, its intended uses and its purpose.

Facial recognition is no different, on top of asking who paid for the development the test and learning data should also undergo serious scrutiny in order to ensure that the system is learning and catering for people of all ethnicities, skin tones and physical features.

Discriminating against someone based upon their ethnicity and physical features is a gross and serious disrespect to that person, even when this action is done by a computer.

Whilst computers and software cannot be held accountable the developers can and should. Computing professionals should strive and aim to create technologies that do not discriminate against its users.