The DNA Database: Making faking easier

So, the government has decided that it should keep DNA data for six years on people who have not been convicted of any crime, despite most consultation respondents disagreeing with retention.

Why is this a problem? Is it just that we have some strange out-moded belief that people should be treated as innocent until proven guilty? It’s certainly true that the government’s argument is precisely that people who have been arrested are more likely to offend in the future (‘re-offend’ was the revealing term used on radio). We are therefore saying to people that although they have not been convicted, we will treat them as if they had been.

But it’s actually about a lot more than that. Keeping the data on such people does more than just infringe their civil liberties.

The UK has a large and ever-growing database – there are over 5.6 million entries, including innocent people, and it is growing fast; from 2007-9 it went up by 25%. Now, with all this activity and effort, one would assume it was working well, and the police were solving lots more crimes …

But in fact, official records show that over those same two years, the total number of DNA-related detections went down by 25% – hardly a ringing endorsement for the need for even more data to be stored on the database.

But there’s another, even more serious problem. Although the courts believe that DNA data is foolproof, and people have been convicted on DNA data alone, it is in fact possible – even easy – to fake DNA samples.

A recent paper showed that with fairly basic lab equipment, it is possible to fake samples – to create a blood sample using someone’s hair, for example. This is fairly worrying, but at least requires a real DNA sample to start with. But if you have access to someone’s database entry, their DNA fingerprint, then it is possible to create samples to match at will – saliva, blood, whatever. This means that organised crime, for example, can easily plant samples at a crime.

Do we really want an illiberal scheme, that doesn’t seem to help the police much, but does make it easier to frame the innocent?

Julian Huppert is a former Cambridgeshire County Councillor (and Group Leader). He works as a research scientist on on the structure and function of DNA. Julian is currently a member of Liberty’s national council and has more information on the flawed use of DNA on his website.

Read more by or more about .
This entry was posted in Op-eds.


  • There are also well-known problems with DNA evidence used in criminal trials, which come from people misunderstanding the statistics behind the science. New Scientist had a good article on this.

  • Seth Thévoz 16th Nov '09 - 12:59pm

    A very sensible contribution from Julian.

  • Malcolm Todd 16th Nov '09 - 3:16pm

    I don’t like the DNA database, but I struggle to come up with a convincing argument against it that amounts to anything more than “I don’t trust the buggers”. Not a wholly unreasonable standpoint, for sure, but hardly a knockdown. So it’s natural to leap on something like this as evidence that one’s instinct knew what one’s reason couldn’t put its finger on:

    But if you have access to someone’s database entry, their DNA fingerprint, then it is possible to create samples to match at will – saliva, blood, whatever. This means that organised crime, for example, can easily plant samples at a crime.

    If true, this doesn’t just call into question the database, but the whole business of DNA testing. I’ve always been rather suspicious of it, as I am of any sort of evidence which basically depends on a lab and an expert saying “trust me, this is what it means” (see above).

    But of course, distrust works more ways than one. So, having followed through the links to Julian’s lay translation of a scientific paper claiming to show how easy it is to fake DNA results, I heard more alarm bells screeching when I came to this para at the end:

    Lastly, Frumkin et al describe a method they have developed that can discriminate between natural and artificial DNA sequences. They have a company, Nucleix Ltd, that can perform tests on such samples. If this (or another approach) was used as standard in forensic assays, then sample faking in this manner could be prevented. However, this would increase significantly the cost of forensic tests, and is not currently performed.

    Ah. So the evidence for the unreliability of current methods of analysing DNA samples is provided by a team who are trying to flog their new, improved, much more reliable but much more expensive method? Colour me underwhelmed.

    Is there more evidence to back this up out there? Or counter-evidence to the Frumkin (great name) analysis? Or reasons for being less alarmed about it than this article suggests? I’m always as happy as the next person to have my prejudices confirmed, which is a good reason to doubt evidence that seems to do that.

  • A telling statistic is that DNA evidence is only collected at the scene in something like 2% of crimes. Quite why we are wasting money on collecting and storing innocent people’s DNA when either DNA isn’t of practical help or we can’t afford to collect DNA evidence in most crimes I don’t know.

  • Malcolm Todd 17th Nov '09 - 12:00am

    @ Julian
    Thanks, that’s interesting. Do you have any idea how DNA profiling compares with fingerprinting? I mean in terms of the level of accuracy, and in terms of how feasible it is to fake the evidence? The thing is, by and large the state isn’t trying to bang up its citizens for no reason, and would rather find the right people if it can. So using a powerful detection technique is a great idea – providing we can be sure that it’s generally reliable, and providing that there is understanding that it can be faked, which opens up the possibility of defending oneself against the charge if there is contradictory evidence or reason to doubt the credibility of the evidence in a particular case.

    I’m not sure how important that statistic is. It matters which 2% it is. No point spending resources on collecting DNA at the scene of a minor offence like shoplifting, or a crime where the identity of the (alleged) offender is never in dispute.

  • Andrew Suffield 17th Nov '09 - 10:35pm

    Messing with the evidence is dead easy. Here’s the quickest, simplest way: run a hand vacuum over a bus seat, shake the contents out at the crime scene (with thanks to Charlie Stross for the idea). If you want to frame somebody specifically then you’ll need to do their car, but that’s not too hard – any car thief can manage it.

    If you want a better reason to be concerned about DNA databases, consider this: if the BNP got into power, or even just got a few sympathetic people in the civil service, what could they do with this information? You can start with retroactively rewriting immigration laws based on ancestry (easy to identify with a database like that), to exclude naturalised citizens, and go from there.

  • Even without that, as an earlier commenter said, it is the temptation to misuse or misrepresent the results which makes the database dangerous.

    What a court is told is – we have this sample of DNA as evidence. Compared with the defendant’s sample in the database, the match is sufficiently close that only one person in X taken at random from the population would have that close a match, where X is a large number (up to half a million being the largest I have been given).

    The jury, if it is a jury trial, is impressed and left with the impression that there is only a two in a million chance that the DNA does not come from the defendant. Whereas, if we start from the innocent till proved guilty hypothesis, what should be being said is – that still means there are over 100 people in the country that this DNA is as likely to have come from as the defendant. Unless you have compelling other evidence, that is not a sound basis on which to base a conviction.

    Even more significant, note that little phrase “taken at random.” I have had a family law paternity case to adjudicate where the DNA sample was tagged at 1 in 100,000, the man had had sex with the girl at the relevant time, and CSA was asking us to find paternity proved. But in fact, as other evidence showed, the biological father was the uncle of the man before us. They shared specific gene patterns which were rare in the population at large.

    I have no problem with the concept of a DNA test being conducted and introduced as evidence when there is already separate evidence sufficient to justify an arrest on suspicion. I have a huge problem with trawling through a DNA database to find partial matches in the absence of such evidence and then in effect challenging the unlucky person to prove they are not the perpetrator.

  • Excellent post Ian, people don’t actually understand the statistics of DNA databases

  • Malcolm Todd 18th Nov '09 - 12:02am

    I think you’ve nipped the bull squarely in the eye, Ian: it’s the fear of trawling (a la CSI) through a database to find your suspect that would be deeply dangerous. Obviously if there are other, good but not on their own compelling, reasons to suspect someone, then the chance that they are an accidental match becomes exceedingly small. (Though the family argument is an interesting point: it rather puts the defendant in the position of having to incriminate a relative in order to prove his/her innocence. Though I suppose there are other situations where that’s true.) Of course, to the Police the fact that you were arrested once for a completely different crime that they couldn’t prove against you may be considered reason enough. I think I may have found my principled reason to back up my gut instinct. I thank you all!

  • NO one is talking about the people with minor convictions lets help get those people of the police database as well why should they langish on the lawless databases for ever, its not just innocent people who can suffer injustice come on libdems say some thing about the people with minor crim records or are you just using the soft option arguing for the removal of innoent people from the NDNAD to score political points

Post a Comment

Lib Dem Voice welcomes comments from everyone but we ask you to be polite, to be on topic and to be who you say you are. You can read our comments policy in full here. Please respect it and all readers of the site.

To have your photo next to your comment please signup your email address with Gravatar.

Your email is never published. Required fields are marked *

Please complete the name of this site, Liberal Democrat ...?


Recent Comments

  • Elizabeth Pears
    I'm not against UBI but I don't think now is the time for the Lib Dems to be promoting it as a policy. I agree with David Raw's comment about the Tory tabloids....
  • Joe Bourke
    UBI is better termed as a Guaranteed Minimum income and can only be effective in achieving its aims if we are prepared to tackle what Beveridge described as t...
  • Helen Dudden
    I had to pay £2000 for dental treatment as the lockdown closed down NHS dentists. Four broken teeth that were caused health problems. The NHS should have mor...
  • Tristan Ward
    Tory Candidate is one Helen Hurford From Conservative Home: “Hurford is a former headteacher, and has taught in schools across the constituency, before set...
  • nigel hunter
    Onward go the troops!...