The New Frontiers of Violence: When Misogyny Becomes Artificial
The Higher Lens | Special Edition: Women, Power and the Digital Age
What does violence against women look like when it no longer requires physical contact, proximity, or even a real image?
For centuries, women's bodies have been sites of political contestation, social regulation and institutional control. Today, artificial intelligence has introduced another dimension to this history. A woman's face can be extracted from an ordinary photograph, manipulated into sexually explicit material and circulated to thousands of strangers without her knowledge or consent. The image may be fabricated, but the violation is not.
We are entering an era in which technology can manufacture an intimate representation of a person without their participation. Yet the most disturbing aspect of this development is not necessarily the sophistication of artificial intelligence. It is how easily existing forms of misogyny have found expression through it.
AI has not invented violence against women. It has expanded the means through which that violence can be perpetrated, reproduced and normalised.
When Misogyny Becomes Automated
In The New Age of Sexism: How the AI Revolution is Reinventing Misogyny (2025), feminist writer and campaigner Laura Bates examines how emerging technologies risk reproducing the inequalities already embedded within society.
Bates, who founded the Everyday Sexism Project in 2012, has spent more than a decade documenting the experiences of women subjected to harassment, discrimination and abuse. Her latest work moves beyond individual incidents to examine the technological systems increasingly facilitating them.
The distinction is important. Artificial intelligence is frequently presented as neutral: a sophisticated instrument whose moral consequences depend entirely upon its users. Yet technologies are not developed outside society. They reflect decisions about whose experiences matter, which risks deserve attention and what forms of harm are considered acceptable.
Bates's research challenges the assumption that technological advancement necessarily produces social progress. In a 2025 interview with PBS NewsHour, she argued that AI systems risk embedding existing structural inequalities into the foundations of future digital environments.
This raises a fundamental question: if technological progress is built upon existing inequalities, who exactly is progressing?
Consider AI-generated sexual deepfakes. Unlike conventional intimate-image abuse, the perpetrator does not require access to an authentic intimate photograph. An ordinary image, perhaps taken from a public social media account, can become the basis for fabricated sexual material.
The victim may never have participated in the depicted act. She may never have met the person responsible. Nevertheless, she must confront the consequences of an image that appears to represent her.
The violence lies not in whether the image is authentic, but in the deliberate removal of consent and control.
The Political Cost of Being Visible
This question of control extends beyond individual privacy. It reaches into political participation itself.
In October, I had the opportunity to help organise and chair a panel at the Women in Politics Conference at King's College London. The experience encouraged me to reflect on a contradiction at the centre of contemporary debates about women's representation.
We frequently ask how to encourage more women to enter politics, assume leadership positions and participate in public debate. Yet we devote comparatively less attention to the conditions they encounter once they become visible.
Visibility is essential to political representation. It is also a source of vulnerability.
Women in public life have historically faced scrutiny extending far beyond their political ideas: their appearance, clothing, relationships, sexuality and perceived femininity become subjects of public discussion. AI-generated sexual abuse introduces another mechanism through which that scrutiny can be weaponised.
A fabricated intimate image can be used to intimidate a political candidate, undermine a journalist's credibility or discourage an activist from speaking publicly. Its purpose need not be to convince everyone that the image is genuine. Sometimes, humiliation, uncertainty and the threat of circulation are sufficient.
Bates has described receiving sexually explicit deepfake videos depicting herself in retaliation for her writing about misogyny. Her experience illustrates how this technology can be used not merely for sexual exploitation, but to punish women for exercising their voices.
When participation in public life carries the possibility of digitally manufactured sexual humiliation, the implications extend beyond individual victims.
They become questions of democratic equality.
What does equal political participation mean if some individuals must accept forms of intimidation that others rarely encounter?
The Fiction of a Harmless Image
One of the most troubling responses to synthetic sexual abuse is the suggestion that fabricated images are somehow less harmful because the depicted events never occurred.
This reasoning mistakes physical authenticity for the foundation of human dignity.
The harm of sexual violation is not limited to bodily contact. It can arise through coercion, exposure, humiliation and the removal of personal autonomy. A fabricated image may be technologically artificial while producing entirely real psychological, reputational and professional consequences.
There is also a deeper issue of ownership.
In digital environments, a woman's likeness can increasingly be treated as raw material: something to extract, modify, sexualise and redistribute. Her identity becomes detached from her authority over its representation.
This is not simply a problem of misinformation. It is a question of who possesses the power to define another person's body in public.
The term deepfake pornography can itself obscure the nature of the abuse by placing non-consensual sexual exploitation within the language of entertainment. More precise terminology, such as AI-generated intimate-image abuse, makes the absence of consent central rather than incidental.
Language matters because it determines whether society recognises an incident as a technological novelty or a violation of rights.
Innovation Without Accountability
Governments have begun responding to these harms. In the United Kingdom, new criminal offences concerning the creation and requesting of non-consensual purported intimate images came into force in February 2026. Further provisions targeting the production and supply of intimate-image generation tools followed in June.
These developments represent important legal recognition that synthetic imagery can constitute abuse. Yet criminalisation alone cannot resolve the structural conditions enabling it.
The responsibilities of AI developers, platform operators, regulators and law-enforcement institutions remain central to prevention.
Should companies be permitted to release image-generation systems without sufficiently robust safeguards against sexual exploitation? Why should victims bear the burden of identifying, reporting and pursuing the removal of content they never consented to create?
And why does meaningful institutional intervention so frequently appear to follow harm rather than anticipate it?
Technological innovation is often celebrated for its speed. Human rights protections, meanwhile, are treated as regulatory obstacles to be negotiated later.
But innovation cannot be meaningfully described as progress when its costs are disproportionately borne by those excluded from decisions about its development.
The Right to Exist Without Being Recreated
The debate surrounding AI-generated violence against women is ultimately about more than artificial intelligence.
It concerns the relationship between power, representation and autonomy.
It asks whether women can participate in public life without having their identities appropriated, whether technological systems should be accountable for foreseeable harms, and whether society is prepared to recognise violence when it occurs through pixels rather than physical proximity.
My experience at the Women in Politics Conference reinforced the importance of creating spaces where women can speak, lead and influence public discussion. But representation must mean more than securing a seat at the table. It must also mean protecting the conditions that allow women to remain there.
Laura Bates's work reminds us that misogyny does not disappear when society becomes technologically sophisticated. It adapts to the structures made available to it.
The challenge, then, is not merely to make artificial intelligence safer. It is to question the inequalities that determine whose safety is considered expendable in the pursuit of innovation.
A society cannot claim to be advancing women's freedom while developing new mechanisms through which that freedom can be taken away.
The images may be artificial. The power exercised through them is not.
And neither is the violence.
Vlera Gara