As the strapline, “How the AI revolution is reinventing misogyny” suggests, The New Age of Sexism is a forensic examination of the sexism and misogyny built into the latest AI-enabled technology, specifically deepfakes, the metaverse, sex robots, cyber brothels, revenge porn, and AI girlfriends, and in the development and training of AI (artificial intelligence) itself.

As such, the book is required reading for anyone who wants to understand the world everyone, including our children, are increasingly living in. However, I have reservations about some of Bates’s analysis.

In the Introduction, she explains that despite the advances in equality between the sexes in the past century, women and men effectively live in different worlds. Just walking down the same street, their experiences are likely to be fundamentally different – and the same applies online. And she provides some stunning data to illustrate this. Women are 27 times more likely to be harassed online than men are. 50% of women don’t feel safe sharing their thoughts honestly on the internet.

Women using social media have already learnt to self-censor and to disguise their real identities and keep quiet about what they really think as a way of coping with, avoiding, or deflecting, the sexist abuse that is endemic on these platforms – just as they have learnt to take sometimes elaborate precautions ‘in real life’ when travelling alone or to avoid triggering an abusive male partner or colleague, and when navigating the sexual harassment from men and boys that’s rampant in educational settings and many public places. Women already shrink into themselves, stay silent, and avoid eye contact, just to get through their days.

Throughout the book, Bates shows powerfully how AI technology is replicating and amplifying the same sexist and misogynistic attitudes, behaviour and violence that cause such havoc to women and girls on social media and ‘in real life’ – to the point that quite frankly, you are forced to wonder whether humanity as we know it can survive much longer.

Is a society viable in which one half of the population (men and boys) treats the other half (women and girls) as prey to be attacked, silenced, tortured, and ruined? Especially when you consider that society is dependent on women producing and raising new generations of human beings, while the sex class who act as predators are peripheral to this process. (If you don’t believe me, you might want to investigate cattle farming.) Does it not sound like a society, a culture, in terminal decline?

But what disturbs me most about this book is that Bates never seems to ask why. Why is sexism and misogyny so rampant now? Why has it returned with such a vengeance at this historical moment, when we were making real progress on this front from the 1970s up to the early noughties? Why are 12–14-year-old boys now one of the biggest groups of sexual predators? Yes, there were always some children that age who had been abused themselves who would act that out on their younger siblings and neighbours, but they were outliers rather than part of a burgeoning trend as now. What caused this change? Why does Bates never appear to consider the role of online porn and the burgeoning sexploitation industries in all this?

Many older people have a rose-tinted idea of what modern internet porn consists of, thinking of their fathers’ Playboy magazines and titillating photos of topless women. While these might reinforce sexist stereotypes, such photos do not necessarily incite sexual violence. This is not what we are dealing with here. Mainstream internet porn, the kind of porn that kids see with just one or two clicks, is the kind of porn that Andrea Dworkin described in Against The Male Flood:

“Women turned into subhumans, beaver, pussy, body parts, genitals exposed, buttocks, breasts, mouths opened and throats penetrated, covered in semen, pissed on, shitted on, hung from light fixtures, tortured, maimed, bleeding, disembowelled, killed. […]

It is rape and gang rape and anal rape and throat rape: and it is the woman raped, asking for more. […]

It is the conditioning of erection and orgasm in men to the powerlessness of women; our inferiority, humiliation, pain, torment: to us as objects, things, or commodities for use in sex as servants.”

Dworkin wrote this in 1985, when this kind of porn was mostly a secret pursuit of pervy men and not easily accessible by most kids or even many adults. Rapid technological developments since then have changed all that. Now affordable smartphones, high-speed broadband and mobile phone networks, and free tube porn sites have put this kind of porn in everyone’s pockets – accessible with just one or two clicks. Research shows that children are accessing it en masse, often while still in primary school.

Exposing kids to this kind of violent and graphic sexual content is a form of sexual abuse. Early sexual abuse is well understood to lead to hypersexualised behaviour and the acting out of the abuse. We should not be surprised therefore that boys are now acting out the kind of violent sexual behaviour that they have been watching in porn and that girls are being groomed into accepting such behaviour and even thinking it de rigueur.

The impact of heavy internet porn consumption on adults is also well documented. Over time, it can affect the brain and impair decision-making abilities, reduce the ability to curb impulsive behaviour, and increase the likelihood of sexually aggressive and violent behaviour. Webcamming and OnlyFans enable men to order up private online sexual performances – giving them a huge sense of power and entitlement over women at little cost and no risk.

As Kathleen Richardson puts it, porn “functions as a cultural tool to reinforce and maintain the structure of sex inequality, and male domination and female subordination”.

Nowhere does Bates reflect how this and the accompanying explosion of online and in real life prostitution are almost certainly factors in the booming sexism and misogyny that we are currently witnessing and that are getting baked into AI technologies. This feels like a serious weakness in the book, particularly when she comes to discuss solutions. Without addressing the source of the problem, all solutions seem like sticking plasters on syphilitic chancres. Major cognitive dissonance in other words.

Deepfakes

Bates explains that deepfakes are images or videos in which AI technology is used to replace one person’s likeness with another, “making it seem as though a real person has done or said something they didn’t really say or do” [page 12].

She demonstrates how easy it is to create deepfakes by using her own photo in one of the many free and easily accessible apps that specialise in creating porn deepfakes of women. (Unsurprisingly there is almost no demand from women to create such deepfakes of men and most of these apps simply don’t work with photos of men.)

“I feed into an app a press photo of me arriving at an awards ceremony. In less than ten minutes, without spending a penny, I am staring at a highly realistic image of myself standing, completely naked, on a red carpet. The photograph is seamless. It shows ‘my’ breasts, with my hair falling down onto them, and ‘my’ genitals, shaved and exposed. If I didn’t know that it wasn’t my body – not my bikini tan line or my belly button – I would be completely fooled by the image. It looks utterly real.” [Page 13]

Apps and websites offer features, sometimes for a small fee, that integrate a woman’s photo into a variety of porn videos – typically with results that are indistinguishable from reality. She describes how after she published Men Who Hate Women during the Covid lockdowns, she was inundated with online abuse – images of men coming for her with machetes, discussions of the best way to hang her – and deepfakes:

“Of all the forms of abuse I receive, [deepfakes] are the ones that hurt most deeply. The ones that stay with me.” [Page 15]

She explains her shock and the panic on first seeing them, the fear and paranoia about who is responsible. Have her friends and colleagues seen them? What about her boss? Will this become her legacy – all that anyone remembers of her life and achievements when she’s gone – even though it is totally fake?

Bates provides harrowing case histories of how female celebrities, politicians, students and even schoolgirls have been targeted in this way, including in the UK, Australia, Beverly Hills and Spain.

“[T]he abuse and harassment of women ha[s] been turned into a game, complete with incentives and prizes. As a report by social media analytics firm Graphika succinctly put it, deepfake image abuse has shifted ‘from niche pornography discussion forums to a sealed and monetised online business’.” [Page 20]

Just as pornography is about degrading and humiliating women, so is deepfake porn – but with the added advantage that discontented men no longer have to just imagine their uppity female colleague’s face being ejaculated on by multiple men. Through AI’s wizardry, they can now actually watch it happening.

“It is all about putting women in their place. It is all about power and control. It’s not just sexualising them. It’s about subjugating them. […] Silencing them through fear. In another context we might call this terrorism. And if we did that, if we used that accurate language and applied the same political and societal response that we should to any form of terrorism, perhaps we would see more appropriate and urgent action being taken as the technology proliferates.” [Pages 18-19]

The scale of the problem is eye-watering. There are literally thousands of deepfakes of just about every female celebrity and politician – but practically none of their male counterparts. This is assault on the entire female sex class. Every single woman and girl is now a potential target. The consequences for women’s participation in public life, in politics, the law, the music and art world and elsewhere are chilling.

But what is the most common public policy response? Concern about deepfakes being used to put words into (mostly) male politician’s mouths. And that women should protect themselves by not sharing their images! As if that is possible now everyone has a smartphone in their pocket and so much of public life takes place on Zoom.

Schools typically respond inadequately – giving impunity to the perpetrators (invariably boys) and focusing on protecting reputations. There is little or no proactive education in this area. When the apps are so accessible and boys see adults tolerating if not actively using online porn, how are they to know that creating deepfakes of their female class mates is unacceptable? Boys are being groomed for this.

There are no international conventions to deal with this growing phenomenon and there is little legislation to combat it and where there is, it is usually ineffective. England’s Online Safety Act provides weak protection that doesn’t even apply to the small sites, such as university and community sites, where individuals are typically more easily recognised. Furthermore legislation, such as it is, typically focuses on sexualised deepfakes, meaning that deepfakes of women in Iran not wearing the hijab, for example, can lead to terrifying outcomes with no consequences for the perps.

Bates explains that the infrastructure is complex but if there were political will, there are many possible points at which effective action could be taken to ameliorate the situation. For example, search engines could cease returning the deepfake websites and the Apple and Android app stores could cease listing the deepfake apps. There could be robust age verification on the apps. And so on.

But of course, it’s only women and girls! It’s not much of a problem for men, so why should anyone care? After all, the entire porn and prostitution industries are telling us that women and girls are born to be subordinated, silenced, shamed and harassed. [Editor: This is sarcasm.]

The metaverse

If you’re like me, you might have only the vaguest idea of what the metaverse even is. Bates helpfully explains:

“Imagine a sprawling vision of islands and landscapes, buildings and arenas, offices and shops, aquariums and bars, where millions of avatars can stroll through town squares, listen to comedy sets, attend massive virtual concerts, enter boardrooms and meetings with work colleagues, learn in virtual lecture halls (where everything from Roman Empire architecture to surgical procedures will come to life in front of them in 3D) and – perhaps most importantly from the perspective of Meta’s shareholders – buy everything from virtual clothes and accessories to access to exclusive spaces and events. There’s the possibility of virtual towns and cities, where users could build and decorate their own virtual homes – in a sort of Minecraft / Sims mash-up – and then invite friends over to visit.” [Page 59]

She explains that accessing the metaverse requires forking out for virtual reality headsets and additional wearable touch controllers. Rather than interacting using a keyboard or touchpad, this allows you to use your own body movements to control your avatar – which makes everything feel real in a way that traditional video games and apps don’t and leads to a much greater blurring of the boundaries between the virtual and real worlds.

Bates leads us through her own practical excursions into the metaverse, bringing home its immersive and hyper-realistic nature. How her bodily reactions on finding herself in dangerous situations mirror what she’d feel in real life. She witnesses a sexual assault and is disturbed to find child avatars playing a game that involves using a realistic handgun to shoot other players – in an area that had no effective age verification.

She provides evidence of female users being stalked, harassed, groped and even (virtually) raped, with bystanders often supporting such behaviour – just as in the real world – and the totally inadequate or non-existent responses of the tech companies.

“Watching a video clip of this particular incident is deeply uncomfortable and hammers home exactly why researchers have described being shaken by such experiences. One male avatar repeatedly thrusts his groin at the subject of the abuse, who is trapped in a corner on a bed, while another man extends the vodka bottle in her direction, saying, ‘You’re gonna need more of this, shortly.’ ‘Free show! Free show!’ another male avatar crows.” [Page 69.]

She shows that moderation is weak to non-existent and such behaviour is typically ignored. The onus is on users to use inbuilt tools to block or report the incident, but this is neither intuitive or easy (or even feasible) in the heat of the moment and fast-moving action. As a result, most sexist transgression is ignored, which cumulatively leads to a hostile and intimidating environment for women and girls.

“It is dangerous to dismiss forms of suffering that can have a devastating cumulative impact for women and girls as ‘harmless’, ‘low-level’ or ‘just a bit of fun’. In just the same way that repeated experiences of ‘minor’ street harassment, sexist comments, unwanted advances, catcalls, leering, wolf whistles and more can together gradually restrict women’s access to public spaces, the pattern that will result from allowing such ‘minor’ transgressions to go unremarked upon and unpoliced in the metaverse will presumably have a similarly negative impact on women’s and girls’ usage of virtual reality spaces.’ [Page 80.]

This is a serious concern in light of the expectation that more and more education and employment will soon utilise metaverse technology.

Bates also looks at the disturbing reality of perpetrators grooming children, and discusses whether the argument that allowing people to ‘blow off steam’ virtually will improve their behaviour offline. The evidence suggests it doesn’t but rather that it legitimises and normalises such behaviour.

She has many excellent suggestions for improvements but whether any of them will materialise when the big technology companies are driven by profits remains to be seen. Once more, this only or mainly affects women and children, so who cares! [More sarcasm.]

Sex robots

For anyone unfamiliar with what are commonly called sex robots (or sex dolls) and their growing popularity with – you’ve guessed it – male users, the chapter on sex robots would undoubtedly be shocking. She quotes marketing material:

“We’ve recreated the female form so that we can use it how and when we want. We are no longer slaves to the whims, mind games and control of those who would harness our drives for their own. We can finally live as we want … We can now replace women at the drop of a hat.” [Page 118]

Imagine! A fuckable ‘woman’ who never nags; never complains; whose orifices are always moist and welcoming; who can be put on silent mode; who doesn’t have a mind of her own!

Bates interrogates and thoroughly debunks the much-promoted idea that sex robots are a therapeutic solution to men’s loneliness and social ineptitude. Rather they reinforce and normalise the dehumanisation and sexual objectification of women. She makes it clear that as a society it is beyond time that we faced up to the problems of men’s loneliness and social isolation honestly and that we came up with real concrete ethical and sustainable solutions.

She shows how sex robots reproduce heteronormative, ageist, hypersexualised and racist stereotypes of women and girls.

“[A] page offering Black sex dolls and robots opens with the words: ‘They call us black wild beasts.’ As well as racism, colourism plays a major role here: while there are exotified, hypersexualised Latin American sex dolls for sale and others playing into submissive and racist stereotypes of Japanese women, there are almost no dolls available with very dark skins.” [Page 105]

This is deeply disturbing, but there is no mention of how the entire sexploitation industry has long fetishised and commodified women in exactly the same ways. Gail Dines has written on this aspect of mainstream porn extensively, explaining that it constructs, articulates and exploits racial conflict as a “way to enhance the sexual debasement of women”.

Huschke Mau has often spoken about this aspect of prostitution:

“What we see is that prostitution is a very racist system. It is usually the racially discriminated women who enter prostitution – like Roma women from Romania. And prostitution itself is racist too because it fetishizes ethnicity. In Germany we have brothels that have a kind of apartheid system when it comes to the women. You go to the first floor for the Romanian women. You go to the second floor for Asian women. You go to the third floor for African women. We see that prostitution in Germany makes sex buyers more racist. They use very racist and sexist slurs against women and they try to offer refugee women from Syria money for sex.”

Bates does not explore this connection. Readers are given the impression that sex robots are opening up a whole new racist and sexist world – rather than being, as Kathleen Richardson explains, an extension of the porn and prostitution industries. A development, that given the ubiquity and acceptance of these industries, seems practically inevitable.

Bates rightly says:

“Providing men with robots to abuse would only contribute to the normalisation of the dehumanising sexual objectification of women that lies at the root of our day-to-day experiences of misogyny and violence.” [Page 109]

But she doesn’t go on to make the connection that exactly the same thing applies to prostitution, which provides men with real women they can pay to sexually use and abuse.

What’s stopping her? Is she terrified that the liberal feminists and the ‘sex work is real work’ crowd would come after her for crimes of sex negativity and kink shaming? Would her publisher object? Who knows? But it feels like a terrible wasted opportunity.

Cyber brothels and AI girlfriends

This review is already too long, so I won’t go into detail of the chapters on cyber brothels and AI girlfriends (I will try to write a separate article about them), except to say that my response to them is much the same as it is for the sex robots chapter. Again, her otherwise excellent critique stops short of exploring the parallels with ‘in real life’ brothels and prostitution, where the ‘girlfriend experience’ is all the rage.

For example, she says:

“[C]yber brothels explicitly exist to erase from the equation any boundaries, opinions or resistance from your partner. […] It is all about making men’s dreams come true at the expense of women’s voices, autonomy or presence. And that is a very dangerous precedent to set.” [Page 138]

But isn’t that exactly why men pay real women in prostitution in all its forms? To act out sex acts on someone he can control, whose needs and desires he can completely overlook? To re-enact practices they’ve seen in porn that their girlfriends and wives won’t put up with? Don’t believe me? Just read some of the reviews that punters leave about the women they pay to use and abuse.

I’m sorry, Laura Bates, that dangerous precedent was set long ago. That horse has long bolted. That is why we don’t believe it will ever be possible to bring about real change in our sexist and misogynistic culture unless and until we address the sexploitation industry in all its forms and end its cultural and political acceptance.

Revenge porn

What used to be called ‘revenge porn’ is now increasingly referred to as ‘image-based sexual abuse’, apparently in an attempt “to move away from the titillating, colloquial term ‘revenge porn’ and to encourage people to recognise the severity of the problem.” [Page 163]

But to my mind, ‘image-based sexual abuse’ is a vanilla term that in no way conveys the devastating consequences of this practice, in which mainly, but not only, ex-partners share intimate images and videos taken (with or without consent) during their relationship.

In this chapter, Bates uses real cases to reveal the devastating consequences and the typically pitiful response from the authorities – much like with deepfakes as discussed earlier.

The consequences include severe PTSD, the silencing of women, making them afraid, effectively kicking large numbers of women off the internet and out of public life. Female victims being made to apologise – for something someone else has done to them! Girls being forced out of school and university. Women losing livelihoods and custody of their children. All the while perpetrators invariably get off scot-free and sometimes are even publicly celebrated.

This is revenge! This is vengeance! Whether the perpetrators admit it or not. It is vengeance against individual women and all women. Now entrepreneurs are jumping in to monetise and industrialise it. Porn is – as Andrea Dworkin said – “Women turned into subhumans, beaver, pussy, body parts, genitals exposed, buttocks, breasts, mouths opened…”

So ‘revenge porn’ seems like an excellent term to me. It does conjure up something of the small mindedness and vileness of the perpetrators and how it is yet another form of terrorism against women.

Perhaps part of the dislike of the ‘revenge porn’ term is that the ‘sex work is real work’ lobbyists have so successfully promoted the idea that porn and prostitution – ‘sex work’ as they call it – is a free choice for individual women, and not an assault on both individual women’s human rights and all women – as women are positioned as subordinate, subhuman, objects for men’s use and abuse, their sexuality used to shame and even ruin them. So, in the general consciousness, porn and prostitution are now seen as benign and critiquing it is seen as a direct attack on ‘sex workers’. Which shows the power of euphemism to confuse everyone and to obscure the awful reality. Perhaps this explains why Bates appears so unable to see the full picture.

Designing AI

Bates explains how AI is already deeply embedded in our day-to-day life, whether we realise it or not, and is increasingly being used to tackle many of the challenges facing humanity. The concern is that it has to be “trained” on vast amounts of preexisting data which it then regurgitates as required.

Of course, the old adage, garbage in, garbage out, applies. The worry is that there is so much bias in that pre-existing data and AI’s inability to recognise that bias. This was demonstrated just recently when Grok praised Hitler and made extreme antisemitic comments.

Caroline Criado Perez’s bestseller, Invisible Women, is a brilliant investigation of the sex bias in supposedly unbiased data of all kinds – product design, safety aids, medical diagnosis and prescribing, and so on. Almost everywhere the data is based on a one-size fits all approach based on men to the detriment of women, with often tragic and fatal consequences. It took a fierce independent feminist to uncover the scale of this bias whereas AI trained on the data would accept it as fact and repeat and amplify that bias.

We can think of many other areas of bias. We have written about how most institutions in the UK have been captured by lobbyists for the sex industry. As a result, academics who do not agree with the ‘sex work is real work’ ideology typically find it hard to get appointments, or funding for projects that take a different view. Therefore, most academic studies on the sex industry promote the full decriminalisation of the industry. This means that any AI ‘trained’ on these studies would likewise promote a biased and sexist position on this issue and not recognise the inadequacies and errors in many of these studies.

For example, a widely publicised study published in the Journal of Law and Economics recently claimed to have found evidence that liberalising prostitution laws leads to a significant decrease in rape rates. This led to widespread demands for the decriminalisation of prostitution. The claim was based on a comparison of the incidence of reported rape in various European countries that have different prostitution laws. Esther, a prostitution survivor in the NMN group wrote a rebuttal showing that the study is based on a multitude of flawed assumptions and misleading analysis of the data. A data scientist subsequently investigated the statistical analysis and found many other serious issues that call its validity into further doubt. After trying but failing to get the editor to investigate the concerns, they approached Retraction Watch, which published a report. But as the study was published in a highly respected journal, it is likely that AI would accept its findings at face value.

Bates provides many examples of how AI can shape perceptions and amplify inequalities while also generating false information and repeating outright propaganda. She explains that because AI is considered scientific, its results are even less likely to be checked than traditionally created content.

She shows the impact of inherent bias in facial recognition software, AI image generation, recruitment, healthcare, and more. She warns of the particular dangers of training AI on social media and similar content that we all know is rife with sexist and racist content. Once again, there is a lack of political will to properly design and test systems to reduce and eliminate such bias.

Solutions

In the final chapter, on solutions, Bates recognises that there is no magic bullet that will solve the terrifying issues that she has exposed so brilliantly in the preceding chapters. We need to address the underlying sexism and misogyny. But that is no small task – especially if we allow the sexploitation industry to continue to soar while we stay silent about its terrible impact on women and girls and society at large and while we lead men to believe that there’s nothing wrong with paying to sexually use and abuse women and girls.

That said, she has much to say and many excellent specific recommendations that I truly hope that policy makers and tech companies, developers and entrepreneurs wholeheartedly take note of and rapidly implement.

“To make tangible headway, we will require regulations, which will require political will, which will require public pressure, which will require that we wake up and open our eyes to what’s going on now. We owe it to the children of Almendralejo, to Audre Pott, to the woman whose bloodied and bruised post-operative breasts are non-consensually spread across the internet, to the girl who can’t leave her house because of deepfake pornography, to the children whose social media photos are being shared among criminal gangs, to the girls whose first foray into the metaverse might be so traumatic it becomes their last, to the women who male users of sex robots go on to meet next. We owe it to the future partners of the men creating chatbot girlfriends to abuse, to the women being sexually coerced and exploited in Indian digital sweatshops while tagging the data that drives AI systems, to the single mothers being denied credit by biased algorithms, to the girls too terrified to leave their abusive partner because he might share intimate videos of them – and they fear that nobody will help them if he does.”

Conclusion

Notwithstanding my reservations about Bates stopping short of drawing the connections with the sexploitation industry and the imperative to challenge it and men’s demand for it, I think this is an excellent book and I urge everyone to read it. But particularly parents, teachers, tech bros, VAWG professionals and activists, and men. All men. It really is beyond time that men you open your eyes to the harrowing reality that so many women and girls and other marginalised peoples are enduring and that you become part of the movement for change. A more equal world would be better for everyone.

As Arundhati Roy famously said:

“Another world is not only possible, she is on her way. On a quiet day, I can hear her breathing.”

The New Age of Sexism: How AI and Emerging Technologies Are Reinventing Misogyny by Laura Bates

352 pages | Sourcebooks

---

Further reading

• What the idea of “sex robots” tells us about prostitution
• What’s the problem with sex dolls? A conversation with Kathleen Richardson
• From the woman as object to the object as woman
• What’s wrong with prostitution?
• Working as a receptionist in a legal brothel proved to me that prostitution is anything but a normal job
• What is the Nordic Model?