International

Oxford study reveals explosive growth of deepfake AI tools targeting women

Thursday, 8 May 2025 12:08AM UTC

A University of Oxford study exposes a dramatic surge in AI-powered deepfake generators predominantly targeting women, prompting urgent calls for stronger legislation and technological safeguards amid escalating global misuse and abuse.

The Alarming Surge of Deepfakes: A Call for Urgent Action

Researchers from the Oxford Internet Institute (OII) at the University of Oxford have uncovered a staggering increase in the accessibility of artificial intelligence (AI) tools designed to create deepfake images of identifiable individuals. The study, aptly titled "Deepfakes on Demand," identifies nearly 35,000 such tools available for public download on platforms like Civitai. Since late 2022, these generators have been downloaded around 15 million times, with a striking focus on women as primary targets.

Will Hawkins, a doctoral student at OII and lead author of the study, expressed grave concerns over the findings, noting, “There is an urgent need for more robust technical safeguards, clearer and more proactively enforced platform policies, and new regulatory approaches to address the creation and distribution of these harmful AI models." The researchers caution that their insights may only represent the "tip of the iceberg," as their analysis focuses solely on publicly available models, potentially omitting more insidious practices such as the creation of child sexual abuse material.

Women like Emma Watson, the star of the "Harry Potter" franchise and a current student at Oxford, frequently fall victim to these intimate deepfakes. Alarmingly, the data illuminates that about 96% of these deepfake models target identifiable women, including both globally renowned celebrities and ordinary social media users. This phenomenon not only perpetuates harm against individual women but also highlights the broader implications of misogyny prevalent in digital spaces.

The issue of deepfakes is not confined to individual victimisation. It resonates on a societal level, making headlines around the globe. High-profile individuals such as Martin Lewis have recently spoken out against the risks associated with deepfakes. Lewis revealed that a viewer lost £140,000 to a scam advert employing an AI-generated video of him, stating, “I’ve had enough of this; it’s time something was done.” His remarks underscore the urgent need for protective measures both legally and in terms of technological safeguards.

In the UK, legislative steps are being taken to combat this growing menace. Since April 2023, sharing sexually explicit deepfake images has been classified as a criminal offence under an amendment to the Online Safety Act. Moreover, the UK Government aims to facilitate the prosecution of those creating such images as part of its ongoing Crime and Policing Bill, which is currently at the committee stage. Supportive data from the Revenge Porn Helpline indicates a significant rise in intimate image abuse reports—an increase of 20.9% in 2024 alone.

Yet, the urgency to address deepfakes is not restricted to the UK. The Biden administration in the United States is also advocating for action against the rise of AI-generated sexually explicit imagery. With generative AI tools making the creation of realistic fake images alarmingly easy, there is an emphasis on seeking cooperation from tech companies and financial institutions alike to prevent the creation and dissemination of such harmful content. Advocates suggest that voluntary compliance is not sufficient; rather, comprehensive legislative reforms must also be pursued to provide necessary oversight and enforcement measures.

Globally, countries like South Korea face their own deepfake crises. The nation has seen over 800 investigations into deepfake sex crimes in just the first nine months of 2024, a stark increase from 156 cases reported in 2021. Amidst an environment of deep-seated sexism, women are increasingly turning to social media defiantly while grappling with the risks of being targeted with explicit content. Although South Korea has begun to legislate against the creation and distribution of such images, many victims express feeling unprotected, highlighting the ongoing challenges.

Looking ahead, it is clear that while advancements in AI provide remarkable technological capabilities, they also carry significant risks, particularly for women. Legislation, such as the U.S. Congress's recent "Take It Down Act," aims to criminalise non-consensual deepfake pornography and demands decisive action from social media platforms to remove harmful content expeditiously. Success in combatting these online dangers requires not only legal frameworks but a cultural shift to dismantle the oppressive structures that enable such harassment.

The alarming rise of deepfakes should galvanise public discourse and prompt immediate action across the tech industry and legislative bodies. As the call for regulation intensifies, the focus must remain on protecting victims and preventing the exploitation that deepfakes facilitate, ensuring that technology serves humanity rather than undermining it.

Reference Map:

Paragraph 1 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[2]](https://apnews.com/article/c76c46b48e872cf79ded5430e098e65b)
Paragraph 2 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[3]](https://www.huffingtonpost.es/politica/deepfakes-porno-digital-falso-ia-atacar-mujeres-politicas-activistas-periodistas.html)
Paragraph 3 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[4]](https://www.ft.com/content/9eba22b9-a113-47e5-9a8c-2306abf6ec36)
Paragraph 4 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[5]](https://time.com/6589263/taylor-swift-deepfakes-legal-protections/)
Paragraph 5 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[6]](https://qa.time.com/6308786/nina-jankowicz/)
Paragraph 6 – [[1]](https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss), [[7]](https://time.com/7277746/ai-deepfakes-take-it-down-act-2025/)

Source: Noah Wire Services

More on this

https://www.oxfordmail.co.uk/news/25144023.oxford-university-uncovers-dramatic-rise-deepfakes/?ref=rss - Please view link - unable to able to access data
https://apnews.com/article/c76c46b48e872cf79ded5430e098e65b - The Biden administration is urging the tech industry and financial institutions to tackle the proliferation of AI-generated sexually explicit images, known as deepfakes, that are being used abusively. With generative AI tools making it easy to create realistic fake images, the White House is seeking voluntary cooperation from companies to prevent the creation, dissemination, and monetization of such content. The effort is particularly focused on protecting women, girls, and minors from nonconsensual imagery. The administration calls for action from AI developers, cloud service providers, search engines, and payment processors to restrict access and financial opportunities for sites promoting explicit images of minors. Despite these efforts, there is still a need for legislative action to provide comprehensive oversight and enforcement. The issue has gained prominence with cases involving both celebrities and minors, underscoring the urgent need for industry-wide safeguards.
https://www.huffingtonpost.es/politica/deepfakes-porno-digital-falso-ia-atacar-mujeres-politicas-activistas-periodistas.html - El uso de la Inteligencia Artificial (IA) para crear deepfakes y pornografía digital falsa ha aumentado drásticamente desde 2019. Estos videos, en su mayoría pornográficos, afectan principalmente a mujeres, incluidos políticas, activistas y periodistas. Mujeres como Alexandria Ocasio-Cortez y varias candidatas en Brasil han sido víctimas de esta violencia digital, destinada a humillarlas y desacreditarlas públicamente. Deepfakes refuerzan la objetivación y actúan como una herramienta de represión, silenciando a mujeres y excluyéndolas de la política. Un estudio revela que el 98% de los videos deepfake son pornográficos y el 99% de las víctimas son mujeres. La proliferación de estas prácticas misóginas tiene un impacto significativo en la representación política femenina y perpetúa la desigualdad de género. Expertas recomiendan la urgencia de regulaciones éticas y educativas para enfrentar estos desafíos, y figuras como Ocasio-Cortez abogan por legislación para castigar la creación y distribución de este contenido no consentido.
https://www.ft.com/content/9eba22b9-a113-47e5-9a8c-2306abf6ec36 - Young women in South Korea are facing substantial challenges due to deep-seated sexism and the advancement of technology, leading to new forms of misogynistic crimes. After the spy cam epidemic, where women were filmed without consent in private spaces, deepfakes have emerged as a new threat. Using AI, perpetrators create and share explicit fake videos and images of women. This issue has grown rapidly, with over 800 deepfake sex crimes investigated in the first nine months of 2024 compared to 156 in 2021. Despite South Korea beginning to criminalize the creation and distribution of deepfakes and developing tools to combat them, the problem persists. Many victims feel unprotected, prompting some women to remove themselves from social media to avoid being targeted. Experts argue that merely regulating technology is insufficient without addressing the underlying cultural misogyny.
https://time.com/6589263/taylor-swift-deepfakes-legal-protections/ - The sharing of deepfake pornographic images of Taylor Swift on the social media platform X has underscored the need for enhanced legal protections against deepfakes. Despite efforts by X to remove these images, such content often spreads rapidly, making it difficult to eradicate completely. Carrie Goldberg, a lawyer specializing in tech abuse, highlights the increasing prevalence of deepfakes affecting both celebrities and non-celebrities, including children. While some states have laws against deepfakes, there is no federal legislation to address the issue. Bills like the Preventing Deepfakes of Intimate Images Act and the No AI Fraud Act have been proposed but have made little progress. Experts, including Adam Dodge of Endtab, emphasize that deepfakes are part of a broader problem of online violence, particularly against women. They call for stronger legal frameworks and accountability measures for companies that enable the creation and dissemination of deepfakes.
https://qa.time.com/6308786/nina-jankowicz/ - Nina Jankowicz, an expert in disinformation, was appointed by the Biden Administration in April 2022 to lead the Disinformation Governance Board. However, she quickly became the target of a disinformation campaign herself, facing attacks suggesting she was transgender and infertile. These attacks led to her resignation and the eventual disbandment of the board within three weeks. Jankowicz later discovered deepfake pornography videos of her online, illustrating the adversarial nature of AI towards women. She observed that deepfakes predominantly target women, often those in public fields, to undermine and embarrass them. Jankowicz now advocates for urgent government action against disinformation, online abuse, and deepfake pornography. She founded the Hypatia Project to combat these online harms and highlights the need to address AI's predominantly male-dominated perspective, which perpetuates significant harm to women.
https://time.com/7277746/ai-deepfakes-take-it-down-act-2025/ - On April 28, 2025, the U.S. Congress passed the Take It Down Act, a landmark bipartisan bill targeting the harms of AI-generated deepfake pornography. Spearheaded by Senator Ted Cruz following advocacy from teen victims Elliston Berry and Francesca Mani, the law criminalizes non-consensual deepfake porn and requires social media platforms to remove such content within 48 hours of notification. The bill garnered sweeping bipartisan support, passing the House 409-2, with endorsements from both progressive and conservative groups. It bypasses controversy over Section 230 by using FTC enforcement authority under deceptive trade practices. Tech companies like Meta and Snapchat supported the bill due to its balanced approach. Despite strong backing, civil liberties groups expressed concerns over potential misuse and free speech suppression, especially under the Trump administration’s weakened FTC. Notably, First Lady Melania Trump’s support helped galvanize House passage. Advocates hope this legislation will set the stage for broader child online safety reforms. For survivors like Berry and Mani, the bill represents a crucial step in reclaiming dignity and preventing future victimization.

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The narrative discusses recent developments and legislative actions from 2023 and 2024, indicating relatively up-to-date information. However, the specific study details and technological advancements suggest it might be summarising content from the past few months.

Quotes check

Score: 9

Notes: The quote from Will Hawkins, as the lead author, appears original to this context. The Martin Lewis quote is not found in earlier sources, but it is plausible given his public persona and recent involvement in scams. Without earlier online references, it is assumed these quotes are original.

Source reliability

Score: 7

Notes: The narrative originates from Oxford Mail, which is a regional publication with a good local reputation. The information is supported by references to credible entities like the Oxford Internet Institute and recent legal changes in the UK.

Plausability check

Score: 9

Notes: The rise of deepfakes and legislative actions against them are plausible given current technological advancements and societal trends. The focus on women as targets and the involvement of high-profile figures like Emma Watson and Martin Lewis are consistent with broader societal concerns.

Overall assessment

Verdict (FAIL, OPEN, PASS): PASS

Confidence (LOW, MEDIUM, HIGH): HIGH

Summary: The narrative presents timely and plausible information about the rise of deepfakes, supported by credible sources and recent legal developments. The quotes appear original, and the discussion aligns with current societal concerns.

Deepfakes
Artificial Intelligence
Oxford Internet Institute
Online Safety Act
Digital Abuse