Sex with robots: How should the law

SEX WITH ROBOTS: HOW SHOULD THE LAW RESPOND?

MADI MCCARTHY1 WITH TANIA LEIMAN2

While sex dolls have been available since at least the first half of the twentieth century, technology now enables hyper-realistic, internet-connected, individually programmable, sex robots.3 They can blink, smile and moan. They’re anatomically correct with an almost entirely customisable range of facial and bodily features, and bodies with a humanlike temperature. Marketed as a sex aid, a sex substitute and in some instances, a companion,4 demand for female sex robots far outweighs that for male versions, with male customers significantly outnumbering female customers.

These technological developments, coupled with increasing demand and public concern, suggest Australian policymakers are likely to be confronted with calls for the regulation of sex robots in the future. Balancing competing and complex individual and public interests pose new ethical, regulatory and legal challenges for consideration.

ARGUMENTS OPPOSING SEX ROBOTS

Similar to debates about pornography,5 critics argue sex robots objectify and increase the risk of sexual violence against real women.6 Arguably, hyper-realistic sex robots create and perpetuate unrealistic expectations about sexual subservience or subordinance,7 potentially increasing risks of sexual violence against women.8

Echoing Andrea Dworkin and Catharine MacKinnon’s argument that pornography may desensitise men to sexual violence,9 Kate Darling has argued, in the context of all life-like robots, that ‘there is a danger that treating robots violently might have a negative impact on people’s empathy development.’10 If children vandalise robots that behave in a life-like way, this may influence the way they treat living things.11 Similarly, repeated use of sex robots in a violent manner may encourage ‘undesirable sexual acts or behaviors’.12 Some sex robots can be programmed to reject a user’s sexual advances, mimicking a refusal of consent, a key element of sexual offences in Australia.13 For example, when the Frigid Farrah personality14 of TrueCompanion’s multi-personality sex robot Roxxxy is activated, it is ‘reserved and shy’15 and not ‘too appreciative’16 of a user’s sexual advances.

Robert Sparrow notes that, ‘even if the rape of robots does not succeed in promoting – in the sense of increasing – the rape of women, it exhorts and endorses it.’17 However, he considers it may be misplaced to talk of ‘raping’ a robot that lacks the programmed ability to reject or refuse, given a robot’s inability to feel emotions or consent.18 In Sparrow’s view, this would not to be ‘an act of simulated rape or a representation of rape’19 but simply ‘sex with a robot.’20 This highlights the underlying issue of consent that plagues any discussion of ‘rape’ in the context of robots. After all, if sex robots are not human and cannot feel emotions or consent, is consent actually relevant in the context of human-robot sexual interactions?

ARGUMENTS SUPPORTING USE OF SEX ROBOTS

On the other hand, could sex robots provide positive benefits? Others suggest their use in gender-imbalanced countries may actually relieve societal pressure on women, and provide safer alternatives to prostitution and trafficking.21 Advocates claim benefits can include empowering the older population and persons with disabilities, addressing sex-related anxiety, treating sexual dysfunctions, promoting safer sex and creating a safe place for people who ‘feel insecure about their sexual orientation’.22

Several countries in Asia and the Middle East have ‘significantly fewer females than males.’23 In 2017, approximately 115 males per 100 females were born in China, and 111 males to 100 females born in India.24 The ‘natural’ sex ratio at birth is 105 males to 100 females.25 In the United Arab Emirates and Oman, where there is a large male migrant community, women are outnumbered approximately three to one and four to one respectively.26 Gender imbalances have various and differing consequences on men and women. Most obviously, demand will far outweigh supply of available women in the ‘marriage market.’27 Gender imbalance increases familial or societal pressure to marry and have children, while also placing significant value on a woman’s role as a wife, mother, or daughter-in-law.28 Women may also be at an increased risk of emotional, sexual or physical violence or trafficking. Of the excess men who do not marry or have children, ‘the most affected will be those of lowest socioeconomic status, the most uneducated and with fewer opportunities.’29

In a study of ‘attitudes of sex therapists and physicians toward the therapeutic benefits of sex robots’,30 the top three suggested uses were for physically handicapped persons (65%); to be able to live out certain sexual fantasies (61%); and in isolated environments, such as prisons or space stations (50%).31 The top three suggestions for specific diagnoses and situations in sexual therapy where sex robots could be utilised were for patients with social anxiety (50%); for people who do not have a partner but still want to have a sex life ‘without resorting to prostitution or fleeting acquaintances’32 (50%); and premature ejaculation (47%).33

Any regulation of sex robots will thus require delicate balancing of individual interests in autonomy and privacy in one’s own home and sexual relationships, with broader public interests, such as protecting adults from non-consensual sexual activity and preserving public morality.

EXISTING LEGISLATION - CHILD-LIKE SEX DOLLS

While no Australian legislation yet regulates or prohibits sex robots, regulation of child-like sex dolls has been addressed by the Commonwealth, South Australia and Queensland. These statutory provisions may guide any future regulation of ‘adult’ sex robots.

Under the Customs Act 1901 (Cth) (Customs Act), a child-like sex doll is a tier 2 good prohibited from importation into Australia. For child-like sex dolls not identified at the point of importation or made in Australia, possession is also an offence under the Criminal Code Act 1995 (Cth) (Criminal Code). Both offences carry a heavy maximum penalty, 10 years for an importation offence and 15 years for a possession offence.

In 2019, the Combatting Child Sexual Exploitation Legislation Amendment Act 2019 (Cth) (Child Sexual Exploitation Amendment Act) increased measures to protect children from sexual exploitation.34

The Criminal Code’s new section 273A.1 makes it an offence to possess ‘a doll or other object’35 if it resembles a person, or the body part of a person, who is, or appears to be, under the age of 18, where ‘a reasonable person would consider it likely that the doll or other object is intended to be used by a person to simulate sexual intercourse.’36 The maximum penalty is 15 years’ imprisonment. Definitions of ‘child abuse material’ in s.473.1 of the Criminal Code and s. 233BAB(4) of the Customs Act have been expanded to include child-like sex dolls.37

South Australia followed the Commonwealth’s lead with the Criminal Law Consolidation (Child-Like Sex Dolls Prohibition) Amendment Act 2019 (SA) (CLCA Amendment Act), which commenced on 3 January, 2020.38

Two new offences have been inserted into the Criminal Law Consolidation Act 1935 (SA) (CLCA). Section 63AA makes it an offence to produce or disseminate child-like sex dolls.39 Section 63AAB makes it an offence to possess a child-like sex doll.40 Both offences are punishable by a maximum penalty of 10 years’ imprisonment. The CLCA Amendment Act also defined ‘child-like sex doll’, and amended the definitions of ‘child exploitation material’, ‘disseminate’ and ‘material’ to encompass child-like sex dolls in section 62 of the CLCA. 41 The definition of ‘child exploitation offence’ in the Summary Offences Act 1953 (SA) was also amended.42

In Queensland, the Criminal Code (Child Sexual Offences Reform) and Other Legislation Amendment Act 2020 (QLD) commenced on 15 September, 2020.43 It introduced new offences criminalising ‘the possession, production and supply of anatomically correct, life-like child replicas used for sexual gratification’.44

A doll, robot or other object is defined as a ‘child abuse object’ if:45 • a reasonable adult would consider it to represent, portray or give the predominant impression of a person, or part of a person, under 16 years old,

‘irrespective of whether it has adult characteristics’;46 and • it ‘has been used, or a reasonable adult would consider it is intended for use, in an indecent or sexual context including, for example, engaging in sexual activity.’47

Production or supply of a child abuse object is punishable by a maximum penalty of 20 years’ imprisonment, if the production or supply was for a commercial purpose, or otherwise 14 years’ imprisonment.48 Knowingly possessing a child abuse object carries a maximum penalty of 14 years’ imprisonment.49

No other Australian state or territory has specific provisions dealing with or prohibiting possession of child-like sex dolls. OBSERVATIONS FROM CASE LAW ON CHILD-LIKE SEX DOLLS

Since the inclusion of child-like sex dolls in section 233BAB(4)(g) of the Customs Act, a small, but rising, number of cases have come before the courts.50 Various observations can be made about these cases that may inform future decisions regarding the application of any legislation regulating ‘adult’ sex robots.

First, most charges have been laid under the Customs Act. This is unsurprising, given most child-like sex dolls are identified at point of importation and South Australia and Queensland are currently the only states to have specifically legislated against their possession. However, child-like sex dolls may also be captured by the definition of ‘child abuse material’ in other Australian states and territories, as in R v Miao51 and R v CR52, or under section 273A.1 of the Criminal Code, as in R v Garcia53 .

Second, courts have been careful to remain objective in their assessment of child-like sex doll offences. Justice Berman in R v Miao, quoted in DPP v Lou54, remarked that the District Court of New South Wales was not ‘a Court of morals.’55 Courts have instead focused on evaluating the objective gravity of an offender’s conduct.56

Third, in assessing the objective gravity of child-like sex doll offences, courts have routinely considered such offences to be less serious than possessing child abuse material, despite contrary Parliamentary intention. In R v Garcia, Justice Haesler noted the severity of the 15 year maximum penalty in section 273A.1 of the Criminal Code indicated that Parliament ‘does not distinguish between using such dolls for fantasy acts and many of the acts involving indecent or other actions against real children.’57 Nonetheless, possession of a child-like sex doll was considered not as objectively serious as an act involving actual children.58 Similarly, in DPP v Lou, the defendant’s possession of child abuse

material was considered as representing ‘much more serious offending’59 than his offence of importing a child-like sex doll, which was not at the high end of objective seriousness.60 In DPP v Leverton61 , Justice Harbison remarked that a child-like sex doll was ‘a step removed from that actual depiction of children but it is still a very seriously dangerous item for our community.’62

Fourth, importing or possessing a child-like sex doll has not been considered a victimless crime by the courts, despite the low objective seriousness attributed to the offence. In R v Miao, Justice Berman stated that, while no actual child was harmed by the defendant’s possession of a child-like sex doll, ‘any normalisation of sexual acts committed upon children is capable of increasing the likelihood that children will be abused in that way.’63 Similarly, in DPP v Leverton, Justice Harbison remarked that the offence ‘has the potential to normalise sexual behaviour with children.’64

Lastly, courts have consistently imposed sentences at the lower end of the penalty range for child-like sex doll offences, despite the maximum penalty range of 10 to 15 years. This perhaps reflects a consensus that child-like sex doll offences are low on the scale of objective seriousness. In DPP v Lou, the defendant was sentenced to four months imprisonment with immediate release upon entering into a three year recognisance release order of $3,000.65 In DPP v Leverton, a term of imprisonment was considered inappropriate in the circumstances and instead the defendant was required to enter into a two year recognisance release order of $2,000.66 In R v MB,67 the defendant was sentenced to two years and three months’ imprisonment, to be released immediately upon entering into a recognisance in the amount of $500.68 In R v Miao, the defendant was sentenced to two years and three months’ imprisonment, with a non-parole period of one year and three months.69 Of this, two years was attributable to the defendant’s state offence of possessing child abuse material, which encompassed his possession of a child-like sex doll as well as images and videos of child sexual abuse.70 However, R v Miao is distinguishable from the other cases on the basis that Justice Berman did not outline what proportion of the defendant’s sentence was specifically attributable to possession of a child-like sex doll. WHERE TO FROM HERE?

Sex robots challenge existing conceptions of how humans interact with emerging technologies – and they do so in the most intimate way.

While any regulatory response will need to balance a multitude of interests, ethical questions, and legal challenges, the very real potential for this technology to objectify and promote sexual violence against women suggests action is required sooner rather than later.

However, as the cases on child-like sex dolls suggest, even if sex robots are prohibited in Australia, it is likely that the courts may consider such offences to be less objectively serious than sexual offences against humans, and sentences may be more likely to fall at the lower end of the penalty range, even where maximum penalties are equivalent. B

Endnotes 1 LLBLP(Hons), B Accounting, Associate, LK. 2 LLB, GDLP, GCE(HE), Associate Professor and

Dean of Law, Flinders University. 3 ‘Are We Ready for Robot Sex?’, The Cut (Web

Page) <https://www.thecut.com/2018/05/ sex-robots-realbotix.html#:~:text=The%20 difference%20between%20a%20Real,a%20 head%20%C3%A0%20la%20carte.>. 4 Ibid. 5 Andrea Dworkin and Catharine MacKinnon,

Pornography & Civil Rights A New Day for Women’s

Equality (1988). 6 Robert Sparrow, ‘Robots, Rape and

Representation’ (2017) 9 Int J of Soc Robotics 465, 472; Nicola Döring and Sandra Pöschl, ‘Sex toys, sex dolls, sex robots: Our under-researched bedfellows’ (2018) 27(3) Sexologies 51, 54. 7 Sparrow (n 6) 472; Döring and Pöschl (n 6) 54. 8 Sparrow (n 6) 466. 9 Dworkin & McKinnon (n 5). 10 Kate Darling, ‘“Who’s Johnny?”

Anthropomorphic Framing in Human-Robot

Interaction, Integration, and Policy’ in Patrick

Linn, Keith Abney and Ryan Jenkins (eds), Robot

Ethics 2.0 (Oxford University Press, 2017) 173, 179. 11 Ibid. 12 Ibid. 13 Under section 46(2) of the Criminal Law

Consolidation Act 1935 (SA), ‘a person consents to sexual activity if the person freely and voluntarily agrees to the sexual activity.’ See also Crimes Act 1900 (ACT) s 67; Crimes Act 1900 (NSW) s 61HE;

Criminal Code Act 1983 (NT) s 192; Criminal Code

Act 1899 (QLD) s 348; Criminal Code Act 1924 (TAS) s 2A; Crimes Act 1958 (VIC) s 36; Criminal

Code Compilation Act 1913 (WA) s 221BB. 14 Ibid. 15 Salton, Jeff, ‘Roxxxy the US$7,000 companion/sex robot (NSFW)’, New Atlas (Blog Post, 4 February 2010) <https://newatlas.com/roxxxy-us7000-sexrobot/14063/>. 16 Ibid. 17 Sparrow (n 6) 471. 18 Ibid. 19 Ibid. 20 Ibid. 21 Mei Fong, ‘Sex Dolls Are Replacing China’s

Missing Women’, Foreign Policy (Web Page, 28 September 2017) <https://foreignpolicy. com/2017/09/28/sex-dolls-are-replacing-chinasmissing-women-demographics/>; Ben Kwok, ‘Sex doll firms cash in on China’s gender imbalance’,

Asia Times (Web Page, 8 February 2019) <https:// asiatimes.com/2019/02/sex-doll-business-cashingin-on-gender-imbalance-in-china/>. 22 Fosch-Villaronga, Eduard and Adam Poulsen, ‘Sex care robots’ (2019) 11(1) Paladyn 1.. 23 Ritchie, Hannah and Max Roser, ‘Gender Ratio’,

Our World in Data (online, June 2019) <https:// ourworldindata.org/gender-ratio#citation>. 24 Ibid. 25 Ibid. 26 Ibid. 27 Ibid. 28 Ibid. 29 Ibid. 30 Christiane Eichenberg, Marwa Khamis and

Lisa Hübner, ‘The Attitudes of Therapists and

Physicians on the Use of Sex Robots in Sexual

Therapy: Online Survey and Interview Study’ (2019) 21(8) Journal of Medical Internet Research 1, 1. 31 Ibid 7-8. 32 Ibid 7-9. 33 Ibid. 34 Explanatory Memorandum, Combatting Child

Sexual Exploitation Legislation Amendment Bill 2019 (Cth) 2. 35 Criminal Code Act 1995 (Cth) s 273A.1(a). 36 Ibid s 273A.1(c). 37 Explanatory Memorandum, Combatting Child

Sexual Exploitation Legislation Amendment Bill 2019 (Cth) 3. 38 Criminal Law Consolidation (Child-Like Sex Dolls

Prohibition) Amendment Act 2019 (SA). 39 Criminal Law Consolidation Act 1935 (SA) s 63AA. 40 Ibid s 63AAB. 41 Ibid s 62. 42 Summary Offences Act 1953 (SA) s 74BN(1). 43 Criminal Code (Child Sexual Offences Reform) and Other

Legislation Amendment Act 2020 (QLD). 44 Explanatory Memorandum, Criminal Code (Child

Sexual Offences Reform) and Other Legislation

Amendment Bill 2019 (QLD) 1. 45 Ibid. 46 Ibid. 47 Ibid. 48 Ibid s 228I. 49 Ibid. 50 R v Miao [2016] NSWDC 181; DPP v Leverton [2017] VCC 1652; R v Millan [2018] WADC 110;

DPP v Lou [2019] VCC 1399; R v CR [2019]

NSWDC 884; R v MB [2019] NSWDC 992; R v

Garcia [2020] NSWDC 55. 51 R v Miao [2016] NSWDC 181 52 R v CR [2019] NSWDC 884 53 R v Garcia [2020] NSWDC 55 54 DPP v Lou [2019] VCC 1399 55 R v Miao (n 51) [1]; DPP v Lou (n 54) [14]. 56 Ibid. 57 R v Garcia (n 53) [18]. 58 Ibid [21]. 59 DPP v Lou (n 54) [49]. 60 Ibid [16]. 61 DPP v Leverton [2017] VCC 1652. 62 Ibid [27]. 63 R v Miao (n 51) [18]. 64 DPP v Leverton (n 61) [39]. 65 DPP v Lou (n 54) [58]-[59]. 66 DPP v Leverton (n 61) [47]-[51]. 67 R v MB [2019] NSWDC 992. 68 Ibid [13]. 69 R v Miao (n 51) [26]. 70 Ibid.