There is a lot of
unease about the way in which the government and police forces are extending
the use of Live
Facial Recognition technology, using it at a range of events to pick out
people who are suspected of committing an unrelated crime in unrelated
circumstances. There is a lot of good evidence to suggest that, in its current
state, the technology is prone to bias: having largely been trained on white
male faces, for instance, it is less effective at correctly identifying faces
of a different colour or gender. There’s no reason why, with time and effort,
using a wider range of faces to retrain the software, that bias can’t be overcome;
on a rational basis, the undoubted bias is more a reason for caution in rushing to
apply the technology than for rejecting it.
Proponents claim
that no law-abiding citizen has anything to fear from the technology in itself;
it’s merely an example of police forces using the latest techniques to improve
their record of apprehending criminals and suspected criminals. Again, it’s a
rational position to take: who wouldn’t want the police to use the best and
most effective methods and technology to reduce crime and its impact?
None of that removes
the unease, however. Part of that is a fear of how the data could be misused:
recording who was at which event on a giant database could enable authorities
to put together a lot of information about the movements of individuals who have
never been suspected of anything. It doesn’t have to be used that way, of
course – but there is a natural and, on the basis of experience, not wholly
unreasonable lack of trust about governments and police forces. Even if there
are rules and regulations governing the recording and retention of data about
the movements of individuals who are, at the time, of no interest to the
police, how certain do we feel that the authorities will keep to those rules?
And never be tempted to trawl back through old recordings and images?
The underlying issue
is about the relationship between the citizen and the state. Theoretically, the
principle of democracy is that the state is there to serve the citizens, but
the practical truth is that the state has always served, primarily, the
interests which control it. That was bad enough when the capacity of the state
to control and surveil was limited by the capacity of the state’s forces,
primarily manpower. But when backed up by what is, for a finite population
within a defined geographical area, a virtually limitless capacity to store
data, the balance of power shifts significantly. With both the traditional ‘main’
parties having fallen under authoritarian leaders, who are becoming more so in
the belief that that is what Farage’s prospective voters want, the idea of the
state as the protector of citizens, rather than the controller of walking
economic resources, is fast receding.
The question we face
is a simple one: in the interests of preventing crime and apprehending those
who commit it, how much extra power to monitor and record the movements and
activities of all of us – criminals and non-criminals alike – are we prepared
to cede to the state? I fear that the trade-off
isn’t being well understood, but if we don’t think about it now, it gets harder
to see how any new use of technology will ever be rolled back at some future
point.