The Burden of Proof

Last updated on: 24 April 2012

As you can see from the results of the Global Criminalisation Scan, there are a multitude of laws governing HIV prosecutions.

This article will focus on prosecutions for sexual HIV exposure and HIV transmission, which appear to be the most common types of prosecution worldwide.

To make things clearer, there are generally two kinds of laws:

  • HIV exposure laws – these penalise someone diagnosed positive for HIV for not disclosing their HIV status before having sex that in theory – but often not actually – risks transmitting HIV to the other person.
  • HIV transmission laws - which penalise diagnosed HIV-positive people who have not disclosed their HIV status when sexual HIV transmission has apparently occurred, as a result of otherwise consensual sex. 

HIV exposure: he said/she said

Jurisdictions that criminalise HIV exposure do not have to prove very much to bring home a conviction. They must prove that the accused had tested HIV-positive; knew the results of that test; in some jurisdictions, understood that they were infectious and understood that condoms would make their sex ‘safer’; and in some jurisdictions, knew their legal obligation to disclose before any kind of sex that might risk transmission.

Proving that the accused knew these things usually requires a court-ordered disclosure of confidential medical records along with evidence from their doctor or another healthcare provider that had counselled the accused. (We consider the implications of such breaches in confidentiality in detail in Do we have a right to privacy?)

Once that is established, then the whole case becomes about two versions of events; one put forward by the complainant(s), and the other put forward by accused. As a result, many cases before the courts have simply pitted one person’s word against the other. Since the defendant is HIV-positive (and thus stigmatised), one could argue that they are already at a disadvantage, even if the court system claims to be fair and just.

For example, a judge in Canada recently showed himself to be ignorant about elementary facts about HIV transmission risks. In January 2008, he was investigated for misconduct after insisting a witness with HIV wear a mask while testifying in his courtroom.

Another judge, from Australia, has such a strong view on the ‘harm’ of a person with HIV having unprotected sex with an HIV-negative person without having first disclosed their status (even with no transmission), that he announced, in December 2007, that Victoria state’s maximum five-year penalty “is not really an adequate reflection of the seriousness” of the offence.

Yet one has to ask what exactly is the ‘harm’ of HIV exposure? What damage has occurred that it requires someone to be locked up in prison for five years (or, in some jurisdictions, much longer, such as the case of a man in the US state of Arkansas recently jailed for 12 years for not telling his girlfriend he had HIV)?

So, for a defendant with HIV, who may also be a marginalised member of society (possibly with a different skin colour or country of origin to most of the jury), the outcome will depend on which testimony appears more credible to a judge and jury.

One of the ways in which the criminal law appears not to make allowances for variations in human experience is the fact that it treats disclosure of HIV status as if it’s a purely legal contract, like telling the drycleaner what you want done to your trousers. But it’s not so easy. Around the world, HIV remains highly stigmatised, and there are strong links between the stigma and discrimination people with HIV face and their willingness (and ability) to disclose their HIV status to their sexual partner(s).

Besides, disclosure on its own does not mean that HIV exposure or transmission is less likely to occur. (i) 
The real question is what precautions are necessary, and whether they are taken. In addition, these laws do not always recognise that some people with HIV disclose their status to sexual partners in ways that do not include direct and open discussion of HIV. Of course, coded or implicit disclosure (such as leaving HIV medication by the bed) may lead to miscommunication.

Consequently, the ‘proof’ in HIV exposure cases relies primarily on the necessarily imperfect testimony of complainants. (For an example of the way HIV exposure laws can be shown to be arbitrary, with evidence based on gossip and innuendo, see the case of this woman from Ontario, Canada, who in August 2008 was acquitted for allegedly having unprotected sex with a man without disclosing her HIV status.)

HIV transmission: fitting together pieces of puzzle

The first and most important point to make about evidentiary issues in prosecutions for HIV transmission is that if the accused pleads guilty before a trial, then there is no possibility of evidence coming to light proving that they were not guilty. This can only happen during a trial when both sides have to provide full disclosure of evidence relating to the alleged ‘crime’.

Pleading guilty to receive a reduced sentence is not necessarily a wise move – just because someone feels guilty does not necessarily mean that they are. After all, HIV transmission is often a silent event that cannot be predicted, and can rarely be pinpointed to a particular sexual encounter.

A good example showing how individuals may not really know who infected them is a study (ii) that examined (using phylogenetic analysis: see below) whether contact tracing (when an HIV clinic asks a recently diagnosed individual to name their recent partners, so that they too can be tested) was effective. It found that of 41 people named as the source of infection by individuals who had recently tested positive, only 14 could have possibly been the person who may have infected them. This study also gives cause for concern about the risk of false accusations by sexual partners of those with HIV.

Proving that one person infected another is not an easy task, and although prosecutors, judges and juries may think that scientific evidence can prove this with as much certainty as, for example, DNA testing, this is not so.

Criminal HIV transmission cases must use a combination of virological, immunological and epidemiological evidence – usually obtained from blood tests and medical records – to be able to understand the:

  • disputed fact (could the complainant have been infected in another way, or by someone else?)
  • timing (did the alleged criminal HIV transmission take place before or after the accused was diagnosed HIV-positive?)
  • and direction of the alleged criminal transmission (could the complainant in fact have infected the defendant?).

Consequently, the individual virological, immunological and epidemiological elements of the evidence can only be small pieces of a much larger puzzle.

HIV transmission: scientific evidence may be flawed

A common misconception about HIV transmission prosecutions is that if the complainant(s) and defendant share the same HIV subtype (e.g. subtype A, subtype B, subtype C etc.), then the defendant must have infected the complainant(s). However, since even the rarest of HIV subtypes have been found in more than two people, sharing the same HIV subtype does not conclusively prove anything.

Furthermore, each individual can be infected with a variety of related but slightly different genetic versions of the same HIV subtype. These are known as ‘quasispecies’, and can be examined in much greater detail by analysing HIV’s genetic code (RNA), in a process called genotyping.

Genotyping – or to be more specific, comparing different genotypes via a complex computational tool known as phylogenetic analysis to estimate how closely related two or more samples of HIV are likely to be – can be very useful when used to assess trends in HIV transmission on a population level.

However, it is of limited use on an individual level, such as in criminal cases. This has now been recognised in English law, following the acquittal of a gay man for ‘reckless’ HIV transmission in August 2006 (iii).

Phylogenetic analysis examines very small differences in different parts of HIV’s RNA, in order to estimate how these HIV strains are genetically related. However, HIV, unlike human DNA samples or fingerprints, is never unique to an individual, and for a variety of reasons no current method of phylogenetic analysis can be considered 100% accurate (iv).

Importantly, even if two viruses are found to be very closely related, they are seldom – if ever – unique to just two people. In addition, phylogenetic analysis can neither reliably estimate the timing nor the direction of transmission.

Nevertheless, if phylogenetic analysis is carried out rigorously it is reliable enough to show when two viruses are not closely related to each other. In other words, phylogenetic analysis can exonerate a person accused of criminal HIV transmission.

This was made clear for the first time during the testimony of the expert witness for the defence, virologist Dr Anna Maria Geretti of London’s Royal Free Hospital, during a 2006 trial at Kingston Crown Court in London, England. The trial – only the second in England & Wales for ‘reckless’ HIV transmission between two gay men – was held because the complainant believed that the defendant had infected him with HIV, and had done so without disclosing his HIV-positive status before they both agreed to have unprotected sex during their short relationship.

Dr Geretti testified that although the two men's viruses were genetically similar, phylogenetic analysis was unable to answer questions of timing or direction, nor rule out the possibility that a third party with genetically similar virus may have been responsible for both their infections.

Using evidence from the complainant’s sexual health clinic notes, the defence then argued that since the complainant had engaged in high-risk sexual activities with other men between testing HIV-negative in 1999 and HIV-positive in 2004, it was indeed possible that he had been infected by an unknown third party during this period.

During his summing up, Judge Benning felt duty-bound to direct the jury to acquit, since there was a “clear possibility" that a third party had infected the complainant. (v)

Since then, defence lawyers in England & Wales have regularly challenged scientific evidence based on the possibility that individuals other than the defendant may have been responsible for the complainant’s infection. In most cases, judges have dismissed the charges (vi).

Few other jurisdictions seem to understand fully how limited this kind of scientific evidence is, and in many countries, particularly in the global South, phylogenetic analysis – which is costly and time-consuming – may not be available even to help exonerate the accused.

The point is that criminal HIV transmission cannot be proved conclusively by scientific evidence alone, without at least some element of doubt. The amount of the uncertainty will often be considerable and if the defence and their expert witness present it properly it will make a significant legal impact.

i Galletly CL and Pinkerton SD. Conflicting messages: how criminal HIV disclosure laws undermine public health efforts to control the spread of HIV. AIDS Behav. 10: 451-461, 2006.

ii Resik S et al. Limitations to Contact Tracing And Phylogenetic Analysis in Establishing HIV Type 1 Transmission Networks in Cuba. AIDS Research and Human Retroviruses. 23(3): 347-356, 2007.

iii Carter M. Prosecution for reckless HIV transmission in England ends with not guilty verdict. August 9th, 2006.

iv More detailed information about the use and limitations of phyologenetic analysis in a court of law can be found in Bernard EJ et al. HIV Forensics: the use of phylogenetic analysis as evidence in criminal investigation of HIV transmission, February 2007, published by NAM/NAT (available at

v Quoted in O’Connor C. Law and disorder. Positive Nation (126) October 2006.

vi Bernard EJ. Reckless HIV transmission case dismissed due to insufficient evidence. February 11, 2008.



Go back