'Virtual rape': AI and deepfakes are silencing women in public life, UN report
More than 41% of women said they had self-censored on social media to avoid abuse, while 19% had pulled back from speaking out in a professional context.
Artificial intelligence-powered abuse is pushing women out of public life, according to a new report by UN Women.
The major study found that female journalists, activists, and human rights defenders are facing rising online violence, which includes AI-generated deepfakes and what researchers are calling "virtual rape".
The study, Tipping Point: Online Violence Impacts, Manifestations and Redress in the AI Age, was published by UN Women in partnership with researchers at City St George's, University of London, and TheNerve, a digital forensics lab founded by Nobel laureate Maria Ressa.
More than 640 women in public-facing roles from 119 countries were surveyed in late 2025.
The survey found that 27% of women received unsolicited sexual advances or unwanted intimate images, and 12% had their personal images, including those of an intimate nature, shared without their consent. Another 6% of women were subject to deepfakes or manipulated imagery.
The attacks were “often deliberate and coordinated, aiming to silence women in public life while undermining their professional credibility and personal reputations,” the study found.
A main culprit in online abuse is deepfake tools, which use AI to superimpose a person's likeness onto fabricated images or videos, often of a sexual nature. They have become cheaper and faster and can produce in minutes nonconsensual images that can be used for harassment.
Alarmingly, more than 40% of women said they had self-censored on social media to avoid abuse, while 19% had pulled back from speaking out in a professional context.
This also resulted in a heavy psychological toll, with one in four women reporting anxiety or depression and 13% of respondents diagnosed with PTSD.
"AI-assisted 'virtual rape' is now at the fingertips of perpetrators. This phenomenon accelerates the harm from online violence inflicted on women in public life," said Julie Posetti, professor of journalism and chair of the Centre for Journalism and democracy at City St George's, and the report's lead author.
“This violence serves to fuel the reversal of women’s hard-won rights in a climate of rising authoritarianism, democratic backsliding and networked misogyny,” she added.
The report also highlighted widespread failures in institutional responses, with 25% of cases reported but only 15% of police taking legal action.
A further quarter of respondents who went to the police said they were made to feel victim-blamed and were asked questions such as “What did you do to provoke the violence?”. An equal proportion or respondents said officers made them feel responsible for protecting themselves from further harm.\
Pauline Renaud, Lecturer in Journalism at City St George’s, and fellow co-author of the study, said:
“We need more effective education and training of law enforcement and judicial actors to support action in cases of technology-facilitated violence against women and girls,” said Pauline Renaud, Lecturer in Journalism at City St George’s, and fellow co-author of the study.
“This needs to be matched by political will to effectively regulate Big Tech companies that use their outsized financial and political power to undermine progress in this area,” she added.
Academy Awards organisers issued new rules on Friday to clarify that acting and writing must be performed by humans and not artificial intelligence to be eligible for Oscars. The new rules also include changes to the international film category, expanding eligibility to include films that won top awards from prestigious festivals like Cannes, Venice and Toronto.
Issued on: 02/05/2026 -
By: FRANCE 24
01:32
Actors created with artificial intelligence will not be eligible for an Oscar, the Academy of Motion Picture Arts and Sciences said Friday as it launched a crackdown on the use of AI.
New rules include a requirement that only real, live human performers – not their AI avatars – are eligible for the film world's biggest prizes, and screenplays must have been penned by a person, rather than a chatbot.
"In the Acting category, only roles credited in the film's legal billing and demonstrably performed by humans with their consent will be considered eligible," the Academy said.
"In the Writing categories, the rules codify that screenplays must be human-authored to be eligible."
The ruling comes days after an AI version of the late Val Kilmer was unveiled to an audience of cinema owners, a year after the "Top Gun" star's death.
A youthful, digital version of Kilmer appeared in the trailer for archeological action pic "As Deep as the Grave," telling another character: "Don't fear the dead and don't fear me."
The project was created with the enthusiastic support of the actor's family, who granted access to Kilmer's video archives, which were used to recreate the actor at multiple stages of his life.
The use of artificial intelligence remains one of the most sensitive issues in the entertainment industry and was central to the 2023 strikes that shut down Hollywood, as actors and writers warned that unchecked technology threatened their livelihoods.
Other updates to the Academy's rules include a change in the way that films can be nominated for best international feature.
Until this year, only a film selected by an official national grouping could be entered – a problem for any critical movie made in an authoritarian state.
For example, Iranian director Jafar Panahi's "It Was Just an Accident" was nominated earlier this year as a submission from France.
Under the new rules, a non-English language film also can be submitted in the category if it wins a qualifying award at a major international film festival, including Cannes, Berlin, Busan, Venice or Toronto.
In that same category, the film will be deemed the nominee and not the country, and its director will be "listed on the statuette plaque after the film title" along with the country if applicable, the Academy announced.
(FRANCE 24 with AFP)
Anthropic in talks to secure UK-based Fractile AI chips and diversify supply
Copyright AP Photo/Richard Drew
Anthropic is reportedly in discussions to procure high-performance AI chips from Fractile, a London-based start-up. The agreement aims to diversify the hardware supply chain of the Claude creator, namely away from Nvidia.
The major AI company Anthropic is exploring a potential partnership with the British semiconductor firm Fractile to secure a steady supply of chips for custom inference and reduce the significant overheads associated with current semiconductor solutions.
According to reports, these talks represent a strategic effort by the San Francisco-based firm to decrease its dependency on Nvidia whilst enhancing the speed and efficiency of its current and next-generation models.
As the global demand for generative AI capacity continues to climb, the financial burden of the hardware required to run these systems has become a primary hurdle for developers.
Anthropic, which has received multi-billion-dollar investments from both Amazon and Google, currently relies heavily on Nvidia’s H100 units alongside custom processors provided by its cloud partners.
However, the high market price and limited availability of these industry-standard chips have squeezed profit margins, prompting firms to look elsewhere.
According to industry analysts, a deal with a specialised firm like Fractile could allow Anthropic to exert greater control over its technical infrastructure.
This strategy reflects a broader trend among tech giants, including Microsoft and Meta, who are increasingly moving away from general-purpose chips in favour of internal or boutique designs.
The major AI company Anthropic is exploring a potential partnership with the British semiconductor firm Fractile to secure a steady supply of chips for custom inference and reduce the significant overheads associated with current semiconductor solutions.
According to reports, these talks represent a strategic effort by the San Francisco-based firm to decrease its dependency on Nvidia whilst enhancing the speed and efficiency of its current and next-generation models.
As the global demand for generative AI capacity continues to climb, the financial burden of the hardware required to run these systems has become a primary hurdle for developers.
Anthropic, which has received multi-billion-dollar investments from both Amazon and Google, currently relies heavily on Nvidia’s H100 units alongside custom processors provided by its cloud partners.
However, the high market price and limited availability of these industry-standard chips have squeezed profit margins, prompting firms to look elsewhere.
According to industry analysts, a deal with a specialised firm like Fractile could allow Anthropic to exert greater control over its technical infrastructure.
This strategy reflects a broader trend among tech giants, including Microsoft and Meta, who are increasingly moving away from general-purpose chips in favour of internal or boutique designs.
A shift in memory architecture and a boost for British technology
Founded in 2022 by Oxford PhD Walter Goodwin, Fractile has gained significant attention for its unconventional approach to processor design.
Unlike standard chips that must constantly shuttle data between the processor and separate memory modules, Fractile’s "memory-compute fusion" architecture keeps data directly on the chip using static random-access memory, or SRAM, which does not need to be refreshed.
According to the British start-up, this method can run large language models up to a hundred times faster than existing hardware while lowering operational costs by 90%.
While these performance claims are impressive, the technology is still in the development phase.
Fractile has not yet launched a commercial product, and its specialised chips are not expected to be ready for full-scale data centre deployment until 2027.
Despite the long timeline, the start-up is reportedly in negotiations to raise $200 million (€170.5m) in funding at a valuation exceeding $1 billion (€853m).
The potential partnership highlights the growing significance of the UK’s semiconductor sector on the world stage. If a formal agreement is reached, Fractile could become Anthropic’s fourth major chip supplier, joining the ranks of Nvidia, Google and Amazon.
According to market reports, the discussions remain at an early stage and no binding contract has been signed.
However, the interest from a major player such as Anthropic suggests that in the AI race, the ability to deliver faster and cheaper compute power is the defining factor.
Founded in 2022 by Oxford PhD Walter Goodwin, Fractile has gained significant attention for its unconventional approach to processor design.
Unlike standard chips that must constantly shuttle data between the processor and separate memory modules, Fractile’s "memory-compute fusion" architecture keeps data directly on the chip using static random-access memory, or SRAM, which does not need to be refreshed.
According to the British start-up, this method can run large language models up to a hundred times faster than existing hardware while lowering operational costs by 90%.
While these performance claims are impressive, the technology is still in the development phase.
Fractile has not yet launched a commercial product, and its specialised chips are not expected to be ready for full-scale data centre deployment until 2027.
Despite the long timeline, the start-up is reportedly in negotiations to raise $200 million (€170.5m) in funding at a valuation exceeding $1 billion (€853m).
The potential partnership highlights the growing significance of the UK’s semiconductor sector on the world stage. If a formal agreement is reached, Fractile could become Anthropic’s fourth major chip supplier, joining the ranks of Nvidia, Google and Amazon.
According to market reports, the discussions remain at an early stage and no binding contract has been signed.
However, the interest from a major player such as Anthropic suggests that in the AI race, the ability to deliver faster and cheaper compute power is the defining factor.
How is the EU breaking the 'grip' of tech giants? Ask the Euronews AI chatbot
Copyright AP Photo/ Noah Berger
The Commission's first formal review found the Digital Market Act (DMA) “fit for purpose”. While critics warn against procedural compliance issues, the Parliament asks for smoother implementation. Do you want to know how the DMA regulates online platforms? Ask the Euronews AI chatbot.
On 28 April, the Commission found the Digital Markets Act (DMA) to have "opened up new opportunities for businesses and developers, while giving users more control over their experiences and devices".
With external forces pushing back on the DMA, the Parliament called for its improved enforcement during last week's 27-30 April plenary session.
The Commission backed its review with over 450 contributions from open consultations received between July and November 2025.
93 per cent of Europeans used internet platforms in 2024, a 2025 Eurostat report found. High user volume puts major online platforms in a dominant position, turning them into “digital gatekeepers” between millions of users and the rest of the digital economy.
The term refers to their unlimited power in digital markets, allowing them to impose unfair conditions on end users, such as controlling data and influencing competition.
The Commission labelled Alphabet, Amazon, Apple, ByteDance, Meta, Microsoft, and Booking as gatekeepers and classified 23 of their online services as gateways.
Since November 2022, the DMA aims to limit the power of big online platforms to make digital markets fairer, more competitive, and open to all innovators, businesses, and new market players.
It imposes preliminary obligations and prohibitions on large online platforms, rather than applying standard EU competition law.
Do you want to know how the DMA contributes to a fairer online environment? Ask the Euronews AI chatbot!
On 28 April, the Commission found the Digital Markets Act (DMA) to have "opened up new opportunities for businesses and developers, while giving users more control over their experiences and devices".
With external forces pushing back on the DMA, the Parliament called for its improved enforcement during last week's 27-30 April plenary session.
The Commission backed its review with over 450 contributions from open consultations received between July and November 2025.
93 per cent of Europeans used internet platforms in 2024, a 2025 Eurostat report found. High user volume puts major online platforms in a dominant position, turning them into “digital gatekeepers” between millions of users and the rest of the digital economy.
The term refers to their unlimited power in digital markets, allowing them to impose unfair conditions on end users, such as controlling data and influencing competition.
The Commission labelled Alphabet, Amazon, Apple, ByteDance, Meta, Microsoft, and Booking as gatekeepers and classified 23 of their online services as gateways.
Since November 2022, the DMA aims to limit the power of big online platforms to make digital markets fairer, more competitive, and open to all innovators, businesses, and new market players.
It imposes preliminary obligations and prohibitions on large online platforms, rather than applying standard EU competition law.
Do you want to know how the DMA contributes to a fairer online environment? Ask the Euronews AI chatbot!
What is Mark Zuckerberg’s Biohub and can it build AI models of human cells to cure all disease?
Copyright Canva
Mark Zuckerberg and Priscilla Chan’s Biohub is investing $500 million to build AI models of human cells, as tech giants race to bring artificial intelligence into biology.
Meta’s CEO Mark Zuckerberg and his wife Priscilla Chan are building artificial intelligence (AI) models of human cells, in a project they say could help “accelerate the cure and prevention of all diseases”.
Their non-profit, Biohub, last week announced a five-year initiative to create the technologies and datasets needed to build predictive models of human cells.
The organisation said the data it generates will be made open and freely available to researchers worldwide.
Biohub says AI simulations of human cells could allow researchers to study disease digitally at a scale and speed not possible in the laboratory today. If accurate enough, such models could help scientists understand how cells behave in health and disease, reveal the causes of disease and point towards new treatments.
Meta’s CEO Mark Zuckerberg and his wife Priscilla Chan are building artificial intelligence (AI) models of human cells, in a project they say could help “accelerate the cure and prevention of all diseases”.
Their non-profit, Biohub, last week announced a five-year initiative to create the technologies and datasets needed to build predictive models of human cells.
The organisation said the data it generates will be made open and freely available to researchers worldwide.
Biohub says AI simulations of human cells could allow researchers to study disease digitally at a scale and speed not possible in the laboratory today. If accurate enough, such models could help scientists understand how cells behave in health and disease, reveal the causes of disease and point towards new treatments.
What is Biohub?
Biohub's long-term goal is to cure all human disease through the intersection of AI and biology, Zuckerberg said last year.
In 2016, the couple set up the organisation to bring together scientists and engineers to develop technologies that “observe, measure and program biology at the cellular level”.
Biohub said it has since gathered the largest single-cell datasets globally and built specialised large-scale computing infrastructure dedicated to biological research.
The new initiative reflects a growing belief across the life sciences industry that AI models trained on vast biological datasets could transform how drugs, treatments and therapies are discovered.
The organisation will spend $400 million (around €348 million) on its own work and make a further $100 million (around €87 million) available to external researchers. Its partners include chipmaker Nvidia and leading research institutions.
Biohub's long-term goal is to cure all human disease through the intersection of AI and biology, Zuckerberg said last year.
In 2016, the couple set up the organisation to bring together scientists and engineers to develop technologies that “observe, measure and program biology at the cellular level”.
Biohub said it has since gathered the largest single-cell datasets globally and built specialised large-scale computing infrastructure dedicated to biological research.
The new initiative reflects a growing belief across the life sciences industry that AI models trained on vast biological datasets could transform how drugs, treatments and therapies are discovered.
The organisation will spend $400 million (around €348 million) on its own work and make a further $100 million (around €87 million) available to external researchers. Its partners include chipmaker Nvidia and leading research institutions.
Data is the challenge
Biohub says scale will be central to the effort as AI predictions become more useful as the volume and quality of biological data increase.
“To build artificial intelligence that can accurately represent the full complexity of biology and accelerate scientific research, we need orders of magnitude more data than exists today,” Alex Rives, Biohub’s head of science, said in an announcement.
“We need new technologies to observe the cell, from the molecular to the tissue level, and in the context of health and disease,” he added.”
But researchers l do not know how much data will be needed to make cellular models accurate enough to produce reliable predictions.
Biohub also said a much greater global effort will be needed to reach the necessary scale.
Rives said he hoped other funders would add to the funding Biohub is making available to outside researchers.
AI-powered biology is an emerging industry, as research organisations, technology companies and drug developers look for ways to use machine learning to understand disease and design new treatments more quickly.
Other technology companies are also pushing into AI-powered biology.
Isomorphic Labs, an Alphabet company built on Google’s DeepMind, is using AI for drug discovery and says it is working to design new medicines.
Microsoft has also released several healthcare AI models, including those for medical imaging, genomics, clinical records and biomedical research, while Nvidia’s BioNeMo platform is being used by life sciences companies for AI-driven drug discovery.
Biohub says scale will be central to the effort as AI predictions become more useful as the volume and quality of biological data increase.
“To build artificial intelligence that can accurately represent the full complexity of biology and accelerate scientific research, we need orders of magnitude more data than exists today,” Alex Rives, Biohub’s head of science, said in an announcement.
“We need new technologies to observe the cell, from the molecular to the tissue level, and in the context of health and disease,” he added.”
But researchers l do not know how much data will be needed to make cellular models accurate enough to produce reliable predictions.
Biohub also said a much greater global effort will be needed to reach the necessary scale.
Rives said he hoped other funders would add to the funding Biohub is making available to outside researchers.
AI-powered biology is an emerging industry, as research organisations, technology companies and drug developers look for ways to use machine learning to understand disease and design new treatments more quickly.
Other technology companies are also pushing into AI-powered biology.
Isomorphic Labs, an Alphabet company built on Google’s DeepMind, is using AI for drug discovery and says it is working to design new medicines.
Microsoft has also released several healthcare AI models, including those for medical imaging, genomics, clinical records and biomedical research, while Nvidia’s BioNeMo platform is being used by life sciences companies for AI-driven drug discovery.
