Would the UK be a happier or more sinister if the police routinely wore body cameras?

The Home Secretary’s suggestion that to save time the police should routinely wear body cameras sparked a rush of publicity, much of it from those opposed to the concept for a variety of reasons, including because it might lead to a fully digitalised criminal justice system  

In my entry for the Leo Anniversary competition in 2001 for papers on “The World and Business Computing in 2051, I summarised expected developments in this area (including the technologies to ensure that the record was indeed unalterable) as follows: “The police electronic notebook, an unalterable record of what the officer or surveillance system saw and heard, will transform the legal process. Lawyers will then create a new world of obfuscation about what it meant.” The technology is supposedly in widespread use in the US and reports to date appear to indicate that it leads serious savings in time and paperwork, better behaviour (by both police and culprits) and more guilty pleas in return for quick trials although the first serious trial in the UK to test whether changes in behaviour against a similar control group who do not use cameras, only started in May. The US claims might also need to be treated sceptically since according to comments in the Wall Street Journal, they think we are ahead and forces like Fergusson had apparently bought the equipment but not deployed it.

Before blogging on the issues I therefore decided to ask the following question of those on the FIPR circulation list: “What are the issues surrounding the use of recordings from police body cameras instead of memories prompted by notebooks written up after the events. This would appear to be a welcome development allowing major savings in police and court time and a reduction in miscarriages of justice. What, if anything, am I missing? What other issues arise?”   

As expected I received some very thoughtful replies and I will take the liberty of reproducing the chain without comment or attribution – although I will be delighted to add attribution is requested.

The first comment was: 

The police and the CPS will have to review the recordings – possibly rather a lot of them in some circumstances. They will all have to be reviewed for disclosure and then disclosed to the defence. The court will need to see them and the defence may challenge any editing to remove unnecessary footage. It’s not clear to me that time will be saved. Some of the recordings may be open to ambiguous interpretations – which could actually lead to the CPS declining to prosecute in circumstances where a crime really has been committed and a notebook record would have secured a conviction. The recordings are likely to contain people other than the defendant – all captured without their informed consent. Where will these recordings end up? Maybe on some reality TV channel. I’m sure there are other issues.

Then came:

One problem will be a loss of the power of discretion. If an officer (and thus their always-on camera) sees something that could be treated as an offence they may feel obliged to arrest and prosecute where without the camera they might choose to handle it in a different way.

For those interacting with officers there could be other effects. You are no longer just talking to a policeman: the camera represents an unknown number of other people with other agendas who can use your words against you later.

We have to assume that in the near future it will be standard practice for all police camera footage to be analysed by computers and stored effectively forever, so everyone within sight/sound of a police officer would be well advised to hide and be silent ‘just in case’…

Are conversations with police officers protected by legal privilege? I suspect not. Suppose you see a fight in the street, run off to find an officer, and gasp out a garbled account: will your confused first impressions be held to be slander against the people involved? (or does
the recording make it libel?)

That contribution led to the comment: 

That suggests such recordings should only be used “without prejudice

To which the reply was:

I don’t really know what that would mean in this case.

I would have thought that the recordings should be usable in evidence, particularly if the recording devices and procedures are carefully designed. It may be necessary for the officer to appear in court to support and validate the recording.

As with all surveillance technologies, the technology itself is neutral but the data can be used for good or ill. I have no problem with police body-worn cameras provided there are strong rules around the use of the recordings. I would want the technology to support the officer in their duty and to provide accountability if they misuse their position, but I would not want it to become a big-brother-style intrusion that affects their behaviour in other ways.

Here are a few ideas. What I am aiming for here is the equivalent of a perfectly-honest policeman with a perfect photographic memory and good drawing ability, and I don’t think anyone could really object to that!

1)    Video and sound recording to run at all times while the officer is on duty. We don’t want any ‘conveniently lost’ recordingsso maybe the devices should just run constantly.

2)    Recordings to be stored securely for a defined period of time, then to be deleted unless required as evidence for a specific case.

3)    Recordings may *only* be accessed in defined circumstances:

* By the officer in person, in order to make reports, support a case, or defend themselves or others.

* By an officer investigating a specific complaint or incident that the recording officer was involved in. The system should constrain this access by time and/or location of recording.

Other constraints may be needed here to avoid creeping surveillance.

* Where a recording shows a person who is charged with an offence, that person’s legal representatives should be given the relevant part of the recording plus enough before-and-after data for them to be certain that they have the full story.

4)    There shall not be any routine trawling or analysis of recordings.

5)    Recording devices to be tamper-resistant, to record GPS time and location data at 1-minute intervals, to encrypt all data such that it can only be recovered using keys held elsewhere, and to sign all data so that the identity of the device can be reliably determined.

6)    Officers to ensure that the recording includes their own face, voice, and shoulder number at least at the start and end of the shift.

7)    Recording format to be designed so that parts of the recording can be extracted for use as evidence without losing the security and authenticity features mentioned above.

Other things could be added, such as recording the IDs of other such recorders nearby.

The final comment was

We should remember th at the purpose of these devices is twofold: to protect the public from the brutal or dishonest officer, and to protect the officer from malicious allegations.  There is therefore a significant benefit of having the device working.

XXXX’s points are good, but the first is not yet viable with the battery power readily available, so it may be better to state that the onus of ensuring the recorder is running at the appropriate time is on the officer, and any investigation into her/his conduct when the device is not running will lead to the investigator drawing the appropriate inference. However there will then (either in this case, or the always on scenario suggested by XXXX) be the problem that officers will hold back if they suspect that their device has failed for any reason.  This may lead to some unfortunate consequences and needs to be thought through carefully.

Most of the above points already appear to have been addressed in the US, where their claims and ambitions are more ambitious . My personal conclusion is that the more modest objective of linking the policeman’s notebook to annotated video logs is most definitely “an idea whose time has come”, whether the motive is to save police and court time or to improve justice. The knowledge that the activities are being recorded might even help reduce (or at least displace) crime. The issue of ensuring the digital record is any more trustworthy than a policeman’s notebook is, however, non trivial. And what about private video logs.

I should perhaps add that for my 2001 paper I assumed the use of secure analogue worms (write once, read many) after the world had lost faith in anything digital.