Barrister Jessica Sobey discusses the Home Office probe launched after it was found that faulty mobile phone data was believed to have been used as evidence in criminal cases for over a year.
Jessica’s article was published in Law360, 8 April 2024, and can be found here.
The Home Office has announced that it is launching a probe after faulty mobile phone data was used as evidence in criminal cases for over a year, as revealed by the Evening Standard on 27 February. The yet unrectified problems, dating back to 2022, concern data provided by the telephone network provider . This article will address computer generated evidence and its use in criminal proceedings, some of the systemic issues arising from its current use, and the importance of allowing the defence to test the validity of the data.
While the Horizon Inquiry led to much needed scrutiny of the confidence placed in computer generated material and the presumption that all such material is reliable, this Home Office probe has highlighted the further need for scrutiny on the presumed reliability of computer-based systems.
In 1999 the Law Commission recommended the repeal of s.69 of Police and Criminal Evidence Act (PACE) 1984. When it came into effect in 2000, the repeal resulted in a common law presumption that computer evidence would be considered reliable unless there was evidence to the contrary.
There was then a shift of the evidential burden, from a position where computer evidence should be subject to proof that it was operating properly, to circumstances where the accused were tasked with rebutting a presumption that it did.
While there are of course instances where the presumption operates effectively to streamline proceedings and reduce unnecessary costs, a far more nuanced approach is required to ensure that evidence is properly examined where it needs to be, so that mistakes such as those made in the Horizon do not occur again.
In the current legal climate, where increasing reliance is placed on computer generated evidence in criminal prosecutions, the justice system finds itself in an Alice in Wonderland scenario where evidence being presented as reliable, is in fact fundamentally flawed. These are not scenarios where defence lawyers have been able to successfully prevent reliance on such material prior to conviction.
These are cases that have already disappeared down the rabbit hole, where the unreliability of data is only coming to light after conviction. This can be seen in the case of Seema Misra[2], a sub-postmistress who was wrongly convicted and jailed while pregnant. While emails revealed that Post Office lawyers had a report on a Horizon bug creating shortfalls at 40 branches in October 2010, this was not disclosed to the defence at Misra’s trial, leading to her conviction. Her conviction was not quashed until 2021.
In an era of almost addictive mobile phone use, the majority of criminal cases now involve communications data, with some of the most serious offences before the criminal courts relying on circumstantial evidence, such as cell site, to demonstrate an individual’s presence at a particular location at a crucial time. They do that by relying on call data records produced by the mobile phone provider. Call data records that include a mobile phone’s Unique International Mobile Equipment Identity (IMEI).
It is understood by the Home Office and the Evening Standard that it is the IMEI that has been affected by the O2 bug; the potential harm being that inaccurate and misleading information could have been placed before juries in criminal trials — juries that potentially relied on communications data to reach a guilty verdict for the accused.
Whilst the Home Office stated that they were aware of the issue with a telecommunications operator and were working together with the relevant organisations, the extent of the problem is unclear, as is the wider impact on a criminal justice system that is already under the strain of crippling backlogs.
Whilst the Government said it could not speculate on how many trials this could impact while an investigation is ongoing, this is perhaps of little comfort for those awaiting trial; some of whom may be doing so from custody.
The Home Office was keen to make clear that analogies are not drawn between this issue and that which occurred in Denmark in 2019 ,[3] where serious concerns arose in relation to geo location data caused by multiple glitches in the software that converted raw data from phone masts in usable evidence. In that instance, problems included the picture presented of a phone’s location being incomplete, linking to wrong masts and the origins of text messages being recorded incorrectly. A situation that certainly put a different spin on the phrase being in the wrong place at the wrong time!
The Home Office told the Evening Standard that the case in Denmark was a systemic issue within the police systems itself, rather than with an individual phone provider. Nevertheless, the result is that potentially misleading information is being used in criminal prosecutions. This quite rightly raises concerns as to the way in which digital evidence is obtained and deployed.
It would be difficult to find a computer expert willing to say that they have created the perfect software. There are always risks of ‘bugs’ undermining the reliability of the data produced or causing systems to malfunction completely. In a society that places increasing reliance on digital systems, this is arguably even more of a possibility than over 20 years ago when s.69 was repealed. Whilst that repeal assisted in tackling an arguably cumbersome and inefficient process, where the vast majority of prosecutions involve the use of digital material, a more nuanced approach is required.
The software used by telecommunication providers is primarily designed to collect data for billing purposes, not for covert surveillance and monitoring. But that is how the data is used by law enforcement agencies. It is used extensively in trials to demonstrate where people were at a particular time, the direction in which they moved and who they were ‘co-locating’ with.
Communications data in recent times is particularly prevalent in trials involving organised crime groups and serious offences, such as importation of drugs, firearms, and murder. Over recent years, the courts have been flooded with prosecutions brought under the umbrella of Operation Venetic[4], the National Crime Agency led investigation into the use of the encrypted network EncroChat.
The reliability of that data has been the subject of complex, lengthy legal argument, and expert evidence. For years, the defence counsel in these cases have raised concerns surrounding the inability to check the reliability of that data due to French Secrecy laws preventing disclosure of the techniques used obtain it. Concerns have consistently been raised as to the reliance on untested computer-generated material, that forms the basis of criminal prosecutions for the most serious of charges.
Whilst the issue regarding O2 is not one likely to impact the Operation Venetic proceedings, which at this stage seem to pre-date the O2 ‘bug’, further light is undoubtedly shone on serious concerns raised by defence practitioners who find themselves unable to challenge presumptions of reliability.
Whilst it is not yet known at this stage how far reaching the recently announced Home Office probe will be, the need for further scrutiny on the presumed reliability of computer-based systems and the evidence that they produce, cannot be understated.
[1] https://www.standard.co.uk/news/politics/investigation-o2-home-office-police-faulty-data-trials-tim-loughton-b1141695.html.
[2] https://www.bbc.co.uk/news/articles/c2eym6qm6ejo.
[3] https://www.theguardian.com/world/2019/sep/12/denmark-frees-32-inmates-over-flawed-geolocation-revelations.
[4] https://www.nationalcrimeagency.gov.uk/news/operation-venetic.
