August was a slow month, but we got a legal case where our client was accused of tampering with a curfew tag, and I was asked for an expert report on the evidence presented by Serco, the curfew tagging contractor. Many offenders in the UK are released early (or escape prison altogether) on condition that they stay at home from 8pm to 8am and wear an ankle bracelet so their compliance can be monitored. These curfew tags have been used for fourteen years now but are controversial for various reasons; but with the prisons full and 17,500 people on tag at any one time, the objective of policy is to improve the system rather than abolish it.
In this spirit I offer a redacted version of my expert report which may give some insight into the frailty of the system. The logs relating to my defendant’s case showed large numbers of false alarms; some of these had good explanations (such as power cuts) but many didn’t. The overall impression is of an unreliable technology surrounded by chaotic procedures. Of policy concern too is that the tagging contractor not only supplies the tags and the back-end systems, but the call centre and the interface to the court system. What’s more, if you break your curfew, it isn’t the Crown Prosecution Service that takes you before the magistrates, but the contractor – relying on expert evidence from one of its subcontractors. Such closed systems are notoriously vulnerable to groupthink. Anyway, we asked the court for access not just to the tag in the case, but a complete set of tagging equipment for testing, plus system specifications, false alarm statistics and audit reports. The contractor promptly replied that “although we continue to feel that the defendant is in breach of the order, our attention has been drawn to a number of factors that would allow me to properly discontinue proceedings in the public interest.”
The report is published with the consent of my client and her solicitor. Long-time readers of this blog may recall similarities with the case of Jane Badger. If you’re designing systems on whose output someone may have to rely in court, you’d better think hard about how they’ll stand up to hostile review.
Paragraph 32 is suitably damning.
Paragraph 33 looks like an opportunity for a research paper :-). I hope you’re allowed to publish it.
I also recall an anecdote about the flakiness and interdependence of systems in tagging solutions; the SMU reported to the central monitoring point via the defendant’s home telephone, and the defendant simply allowed the ‘phone bill to go overdue such that the ‘phone was cut off and therefore the SMU could not raise alerts. Naturally, this was fixed by a software change in the SMU, where it would periodically ‘phone home to verify connectivity, and an alert would be raised at the monitoring point if the SMU had not called in after a specified time. ISTR this was reported on the Schneierblog, however I’m currently failing to locate it.
Actually, the expert report itself is a very good read. Their systems indeed seem completely unreliable, they probably have very few procedures in place, if any. I would guess their procedures are mostly just verbal, neither documented nor reviewed. From the report it seems that their equipment is probably also quite untested from a truly independent point of view — otherwise they would have probably submitted documentation to that effect. I wouldn’t be very surprised if their equipment is not only faulty but of truly low quality. I am appalled that this whole theater can happen in a developed, truly democratic country. Procedures should be put in place where such companies (and the people behind them) are accountable for the misery they cause by their gross negligence.
As someone else who does IT expert witness work (primarily in the US), I found your report wonderfully restrained and yet quite devastating. Thanks for posting it.
Just a mildly amusing ambiguity in the 2nd sentence above:
“Many offenders in the UK are released early (or escape prison altogether) on condition that they stay at home from 8pm to 8am and wear an ankle bracelet so their compliance can be monitored.”
On my first reading of that I thought: how the hell did they get tagged if they escaped from prison?? I’m thinking now that you are just teasing us with this wording 🙂
Extremely interesting read. However, could you give figures for the following statement.
‘Many offenders in the UK are released early (or escape prison altogether)’.