Jim Bates (forensic expert) 23/05/2005

Prosecution only – not to be disclosed

Expert witness Brian Underhill, Celt Ltd - disclosure pack

The contents of Policy Minute dated 8 of April 2004 must be referred to before taking any actions in respect of this disclosure pack.

Brian Underhill is a forensics computer expert. He has given prosecution evidence in numerous cases involving credit card fraud, the internet and hi- tech crime.

1. This disclosure pack attached consists of a report by Professor Anthony Sammes of Cranfield University.
2. The report was commissioned by the CPS, with the permission of Brian Underhill, following the collapse of a high profile case last year concerning allegations of child pornography downloaded from the Internet. That case could not continue after Mr Underhill conceded that some of his evidential findings and conclusions were erroneous.
3. The purpose of the report was to assess whether the errors made were isolated to that case or of a more fundamental nature. Professor Sammes’ conclusion is that the errors are of a more fundamental nature.
4. Professor Sammes report should not be final at this stage; Mr Underhill has been asked to comment on its findings and has done so. He and CELT dispute some of the findings of professor Sammes. They are preparing a response and a meeting between experts is being arranged. It is possible that Professor Sammes will reassess or revise his conclusions in light of these comments.
5. CPS Headquarters has taken the view that the report is prosecution material that might undermine prosecution case given its potential impact on Mr Underhill’s credibility as an expert. Disclosure of this report should only be done in accordance with usual CPIA principles as set out in the Joint Operational Instructions (JOPI). If disclosure is made it should include the information at paragraph 5 above.
6. The report attached has been edited. The editing covers the following categories:

material that might identify the case which predicated the report and lead to the name of the acquitted defendant;

material that refers to another case;

material that relates to named CPS personnel;

prohibited images of children and irrelevant related contextual material.

8. Following your receipt of the report please keep (Name and contact number removed here) at Police Directorate informed of your actions and case progress.

Police Directorate

CPS HQ.

020 7796 8525

Page 2.  

Prosecution only – not to be disclosed

(part removed by editing or missing) the work carried out by DC Robbins and contain identical results without experiencing any difficulty. This is because sufficient information has been provided by DC Robbins about the processes that he used and the data sets that were involved.

Examination of Witness Statements – Mr Underhill

25.1 I then commenced a detailed examination of the five witness statements produced by Mr Underhill. In order to complete this work I had to produce and examine a clone disk of the AJR/1 image files. This was also required for the examination of the AJR Exhibits produced by DC Robbins (see para 22 above). Full details of the production and examination of this clone disk are given at

Appendix 5.

26. In addition I had to carry out a series of experiments on a working AOL 7.0 system in order to establish precisely how it responded to particular situations. A full report on these experiments and my findings is given at Appendix 6.

27. Finally it was necessary to copy out the experimental AOL systems, the email file sets from the AJR/1 image so that they could be viewed and printed out. This process is reported on at Appendix 8a.

28. Full details of my examination of the five witness statements produced by Mr Underhill are given at Appendix 7.

29. In summary I present the following issues that I consider to be of particular significance in respect of the work carried out by Mr Underhill:

29.1 There are many statements of apparent technical fact which are untrue.

29.2 There are technical deductions that are fallacious.

29.3 The thrust of the statements appears in some cases to be slanted towards a particular finding.

 29.4 There are serious omissions.

29.5 The statements do not easily permit others to corroborate the findings.

29.6 There are too many errors of detail for a professional piece of work.

29.7 Where it is necessary to produce a series of reports or statements, it is essential that each is uniquely identified by a separate identification or exhibit number.

29.8 I find that these witness statements fail to meet almost all the requirements listed at para 13.1 above.

29.9 I would not recommend that reliance be placed on these witness statements and I would find it necessary to verify all the stated findings for myself.