国产av不卡一区二区_欧美xxxx做受欧美_成年人看的毛片_亚洲第一天堂在线观看_亚洲午夜精品久久久中文影院av_8x8ⅹ国产精品一区二区二区_久久精品国产sm调教网站演员_亚洲av综合色区无码一二三区_成人免费激情视频_国产九九九视频

Global EditionASIA 中文雙語Fran?ais
China
Home / China / GBA focus

It's my face, but it's not me

Deepfake technology has caused immense distress and concern to innocent victims and society, with experts and social groups demanding regulations to stop the menace in its tracks and prevent it from becoming a larger trend.Wu Kunling reports from Hong Kong.

By Wu Kunling | HK EDITION | Updated: 2025-12-16 09:09
Share
Share - WeChat

Editor's note: Artificial intelligence is revolutionizing our daily lives and the ways we forge connections, transcending the barriers of time and space. The third article in this AI series examines the misuse of generative AI for nonconsensual deepfake pornography, highlighting legal gaps and the urgent need for reform.

It might all begin with a simple, everyday photo a woman shares of herself online, unaware that doing so would create a nightmare for her.

The posting ignites a barrage of indecent images of her — generated in bulk without her knowledge, saved and silently cataloged on a stranger's drive, waiting to be scrutinized by malicious eyes. The face in the pictures was hers, but the image wasn't.

In July, a male law student at The University of Hong Kong was accused of using a generative artificial intelligence tool to create indecent images of nearly 30 females by grafting their faces — mostly from social media screenshots — onto nude photographs.

Some of the victims, who were the suspect's classmates, revealed online that his computer contained about 700 images, including source photos and generated pornography. Those targeted included his university peers, primary school classmates, former secondary school teachers — some were his close friends and others were people he had met only once.

The victims said legal consultations point to a gap in Hong Kong's existing laws. While nonconsensual sharing of AI-generated intimate imagery is barred, merely creating such content without distributing it isn't a criminal offense, leaving no legal recourse to hold the student criminally liable.

The incident became one of Hong Kong's most prominent face-swapping abuse cases, highlighting how such misuse has evolved into fertile ground for more serious crimes. Yet, the regulatory framework is still wanting.

Another case in the Chinese mainland that also drew much attention involved actress Wen Zhengrong. For over half a year, she was cloned with AI from old footage and used for commercial livestreaming.

Blowing her top online, the actress hit back in comment session — "If you're Wen Zhengrong, then who am I?" — only got instantly blocked. Despite her high public profile, she admits it's hard "to prove that I am me".

Beyond China, face-swapping, notably AI-generated pornography, has emerged as one of the most alarming examples of AI technology misuse worldwide, with no nation seemingly able to emerge unscathed.

A recent survey in the United Kingdom found that 35 percent of 1,700 people aged 16 and above in England and Wales admitted to having viewed sexual and/or intimate deepfake content featuring someone they knew (14 percent) or strangers (21 percent).Conducted by police-commissioned consultancy Crest Advisory, the study also found that one in four respondents felt there's nothing wrong with creating or sharing such content, or were neutral about it.

In August, 48 state attorneys general in the United States urged technology platforms to curb the spread of nonconsensual AI-generated intimate imagery, for example, by restricting searches on "how to make deepfake pornography" and blocking terms like "undress apps" or "nudify apps".

Citing a 2023 report by cybersecurity firm Home Security Heroes, the attorneys general noted that 98 percent of computer-generated fake videos online are deepfake pornography.

"The creation of deepfakes is a grave harm to human decency and dignity, especially among our young people," says Massachusetts Attorney General Andrea Joy Campbell.

Emerging form of sexual violence

"Those photos are fake, but too real," says Doris Chong Tsz-wai, a registered social worker and executive director of RainLily — a Hong Kong-based support group combating sexual violence.

Founded in 2000, RainLily has so far helped more than 300 female sexual violence victims annually, and handled over 2,000 hotline calls with 24-hour one-stop support services, including psychological, medical and legal aid. If commissioned by victims, the organization would reach out to online platforms or coordinators of community groups, seeking the removal of nonconsensual intimate images.

In recent years, RainLily has been swamped with pleas for help from victims targeted by deepfake pornography — seven cases in the 2022-23 service year, and eight and 11 in the following two years. Although the figures were modest, Chong is worried that such acts, with basic disregard for consent and respect, could turn into a troubling trend.

In most of the cases, the victims' faces shown in publicly-shared photos were synthesized with a human body, often with exposed private parts and from unknown sources. The victims tend to be young.According to Chong, nearly all who had sought help from RainLily were in their 20s or 30s. It's also learned from social workers and teachers that nonconsensual photo manipulation occurs in secondary schools although the images produced aren't necessarily indecent.

Compared with survivors of real intimate photos leaking, deepfake victims are often "stunned" when they stumble on those images because "they never took them", says Chong.

For most of the victims seeking help, their top priority is not to lodge a report with the police, but to stop the spread of the images immediately. Although they can seek a court order to delete the images on the grounds of voyeurism, many are either unaware of this option or are too desperate to wait through the legal process.

Chong explains that many deepfake victims don't even have a clue who created the images, or when and how they might have begun circulating. This uncertainty deepens their fear, she says, noting that one victim even avoided all social media out of anxiety.

Given these distinct challenges, RainLily has had to adjust its response to such cases.Chong says this type of sexual violence has presented her team with new difficulties.

Existing laws in Hong Kong are not yet well-suited to addressing such situations — a gap that was exposed in the case involving the HKU student. Chong says she believes legislation shouldn't focus merely on whether indecent content was published or even created or not, but on the need to secure consent before using others' image, regardless of the purpose.She notes that many countries have begun enacting or are considering similar laws.

The more pressing challenge, however, is shifting public perception, Chong says.

As these incidents are relatively new, many people have yet to view them as sexual violence. This occurs not only among the public, but also in some professional assessment processes, which significantly affects how cases are handled. She notes that some victims who turned to social workers or teachers for help were greeted with comments like "it was just a joke".

"Due to such incomplete understanding, victims are told the photos are fake and not to take them seriously, while their terror over the ongoing image circulation and the growing misconceptions about them is completely overlooked," says Chong.

She urged the authorities to closely monitor areas that are not adequately covered by existing laws, update legislation in a timely manner and publicize victims' legal options. Furthermore, public education should be stepped up, especially compulsory sex education, to clarify that creating or sharing AI-generated intimate images constitutes sexual violence.

For victims, Chong advises them to seek professional support if they worry about approaching platforms or individuals alone. "Honor your feelings and don't deny yourself," she would tell them.

Governing the entire chain

The China Internet Network Information Center recently issued a report on generative AI application development in 2025, identifying voice and face-swapping as a top governance priority, ranking ahead of copyright infringement, academic misuse and tech ethics.

Liang Zheng, a professor at Tsinghua University and vice-dean of its Institute for AI International Governance, says tackling the issue is "important, but still difficult".

He notes that, from the start, major countries like China and the US have recognized the negative impact of such technology misuse, initially due to harms like disinformation that threaten public interests. But, as the technology has spread, its misuse has shifted from a societal concern to a personal one, damaging reputations and commercial interests directly.

Nevertheless, Liang says the vast, scattered user base and low technical barriers make it difficult for victims to thwart such abuse or promptly trace its origins and dissemination path. Therefore, a more viable governing approach lies in advancing both technological design and legal frameworks to establish a holistic governance model covering the entire chain — from creation to dissemination.

Within this chain, users must obtain others' personal information in compliance with China's Personal Information Protection Law and social norms, while tech companies, at every stage of the chain, must have both the obligation and capacity to exercise oversight. In this regard, he highlights two technical solutions currently being promoted.

In March, several ministries and departments jointly issued measures on labeling AI-generated content, encouraging service providers to embed invisible markers like digital watermarks in these materials.Unlike conventional watermarks, the digital ones are invisible to the human eye, but machine-readable, as they are inserted in the materials' source code. They help platforms to detect such synthetic content during subsequent upload and dissemination and, as a small consolation for victims, can also provide tangible evidence for potential judicial proceedings.

Industry and academic researchers are also refining a trigger mechanism to automatically identify sensitive material in AI-generated images and videos and take action. Liang notes that similar mechanisms for text are already well established, enabling platforms to block or restrict content containing certain keywords related to pornography or violence. He expects the technology to be effectively adapted for visual media in the near future.

Liang, however, warns that deploying these technologies is still difficult and costly, but expects stakeholders to push forward. "Building security checks earlier into the design phase may limit some product or platform functions, but it prevents many problems that could later spiral out of control," he says.

Legally, Liang says the core of addressing such infringements lies in clearly defining the responsibilities of each party in the chain, with judgments based on their specific actions, motives and the harm caused.For example, when a celebrity's face is swapped into a commercial livestream, the creator, who initiates and profits from the act, should bear primary liability. In other cases, platforms or distributors may also be held chiefly accountable.

"Existing laws provide reference principles, but judges must consider the full context of each case," says Liang. He notes that although China's laws on the rights, duties and obligations for various parties in this chain are more detailed than those in many other countries, legal frameworks will always lag behind technological development as innovation constantly produces unforeseen scenarios.

But still, regulating such misuse doesn't have to rely solely on specialized internet or AI laws. It can and should be supported by civil law provisions protecting portrait, reputation and property rights, forming a comprehensive regulatory system.

As AI-generated text, images and videos flood the internet, the old saying "seeing is believing" is losing its hold, says Liang, pointing out that false, incomplete or misleading information has always existed in the television or internet age.

What's new now, he notes, is that this era places higher demands on us — to use technologies as our tools more responsibly and continually strengthen our general knowledge and digital literacy.

"Technology keeps changing, but the principle remains the same — act responsibly and stay skeptical."

 

Top
BACK TO THE TOP
English
Copyright 1994 - . All rights reserved. The content (including but not limited to text, photo, multimedia information, etc) published in this site belongs to China Daily Information Co (CDIC). Without written authorization from CDIC, such content shall not be republished or used in any form. Note: Browsers with 1024*768 or higher resolution are suggested for this site.
License for publishing multimedia online 0108263

Registration Number: 130349
FOLLOW US
 
精品国内亚洲2022精品成人| 日韩高清欧美高清| 在线观看亚洲精品视频| 亚洲成人综合在线| 亚洲色图第一区| 国产精品久久久久9999吃药| 久久免费精品国产久精品久久久久| 国产麻豆9l精品三级站| 久88久久88久久久| 欧美aⅴ一区二区三区视频| 亚洲欧美不卡| 国产精品五区| 亚洲精品乱码| 亚洲片区在线| 国产情侣久久| 噜噜噜久久亚洲精品国产品小说| 99国产精品99久久久久久粉嫩| 欧美日韩亚洲一区| 欧美日本中文| 激情欧美日韩一区| 激情文学一区| 妖精视频成人观看www| 1024日韩| 亚洲激情婷婷| 国产精品综合| 国产日本精品| 久久婷婷麻豆| 日本视频一区二区| 麻豆一区二区三| 精品中文字幕一区二区| 国产中文一区二区三区| 韩国三级在线一区| 国产69精品一区二区亚洲孕妇| 国产不卡在线视频| av电影天堂一区二区在线| av在线不卡免费看| 久久久综合九色合综国产精品| 国产婷婷一区二区| 国产精品国产精品国产专区不片| 中文字幕一区在线观看| 亚洲精品视频免费看| 亚洲国产欧美在线| 欧美体内谢she精2性欧美| 91久久一区二区| 欧美二区三区的天堂| 日韩欧美一区二区在线视频| 亚洲成av人影院在线观看| 日韩精品中文字幕视频在线| 在线观看日韩av| 久久国产视频网站| 7m第一福利500精品视频| 美女扒开尿口让男人操| 国产亚洲第一的欧洲日产| 91xx在线观看| 99爱免费视频| 在线视频专区| 日本在线视频网| 欧美巨大xxxx做受沙滩| 666av成人影院在线观看| 国产亚洲高清一区| 日韩深夜影院| 91九色精品| 亚洲少妇在线| 国模大尺度一区二区三区| av在线这里只有精品| 国产精品三级av| 亚洲成va人在线观看| 欧美亚洲综合色| 精品国产欧美一区二区| 少妇激情综合网| 97视频在线观看视频免费视频 | 欧美日韩看片| 免费日本黄色| 传媒在线观看| 成人三级网址| 亚洲成人人体| 国产欧美自拍一区| 午夜激情久久| 日韩av在线发布| thepron国产精品| 亚洲三级电影全部在线观看高清| 一本到三区不卡视频| 欧美一区二区精美| 综合网中文字幕| √天堂中文www官网| 福利地址发布页| 亚洲52av| 成入视频在线观看| 亚洲高清在线一区| 国产精品久久久久9999赢消| 日韩精品乱码av一区二区| 白白色亚洲国产精品| 亚洲视频在线观看一区| 欧美吻胸吃奶大尺度电影| 亚洲精品久久久久| 欧美激情国产高清| 欧美成人高清手机在线视频| 日本xxxxwwww| 日本视频在线播放| 成人精品高清在线视频| 久久爱www成人| 日韩午夜av| a级精品国产片在线观看| 亚洲综合清纯丝袜自拍| 6080国产精品一区二区| 日韩亚洲国产中文字幕| 欧美乱xxxxxxxxx| 高清日韩av| 黄色av网站在线播放| 久久91视频| 色偷偷综合网| 久久精品av麻豆的观看方式| 国产精品欧美久久久久一区二区| 在线观看www91| 尤物九九久久国产精品的分类| 男人网站视频| 毛片视频免费| a级片免费在线观看| 97久久亚洲| 亚洲欧洲视频| 久久久久久久久久久久久久久99| 一本色道综合亚洲| 国产亚洲aⅴaaaaaa毛片| 亚洲精品自拍区在线观看| 成年人在线免费观看视频网站| 免费不卡视频| 日韩中文字幕视频网| 极品日韩av| 久久久一区二区三区捆绑**| 欧美色视频在线| 日韩性xxxx爱| 玖玖在线视频| 国产区在线观看| 成人动态视频| 久久看片网站| 亚洲欧美精品午睡沙发| 亚洲福利视频在线| 久草在线免费资源| 欧美13~18sex性hd| se01亚洲视频| 亚洲国产成人精品女人| 99re亚洲国产精品| 欧美在线影院一区二区| 欧美成年人视频网站| 九九热青青草| 99自拍视频在线观看| 老牛国内精品亚洲成av人片| 裸体一区二区| 亚洲人成网站影音先锋播放| 亚洲国产精品久久精品怡红院| 一本之道久久| 亚洲第一视频| 色8久久久久| 国产日韩精品视频一区二区三区| 国产欧美一区二区三区在线看蜜臀| 欧美日韩1234| 欧美亚州一区二区三区| 国产69精品久久久久孕妇| 日韩欧美一区二区三区在线观看| 欧美国产专区| 国产欧美一区二区三区在线看蜜臀| 日韩无一区二区| 久草在线资源福利| 免费在线超碰| youjizz欧美| 毛片av一区二区三区| 亚洲国产cao| 日韩中文字幕久久| 夜夜骑天天操| 原纱央莉成人av片| 黄色综合网站| 日韩毛片一二三区| 国产亚洲欧美日韩一区二区| 美女与牲口做爰视频在线观看| 在线观看三级视频| 日本激情一区| 99精品国产视频| 日韩欧美激情四射| 欧美猛交xxxxx| 午夜免费播放观看在线视频| 天天躁日日躁成人字幕aⅴ| 国产一区二区免费视频| 日本道在线观看一区二区| 性欧美激情精品| 天堂av在线免费观看| 911精品国产| 国产精品一二三区| 欧美剧情片在线观看| 亚洲精品久久久成人| 幼a在线观看| 成人在线免费小视频| 久久女同互慰一区二区三区| 精品国产3级a| 五月天丁香在线| a国产在线视频| 欧美日韩 国产精品| 亚洲欧美aⅴ...| 北条麻妃在线一区二区| 天堂影视av| 日韩免费一级| 国产福利一区二区三区视频在线 | 国产高清在线观看| 亚洲精品推荐| 久久无码av三级| 亚洲精品之草原avav久久| chinesemodel无套啪啪| 成人做爰免费视频免费看| 老牛嫩草一区二区三区日本 | 日本道在线观看一区二区| luxu259在线中文字幕| wwwxxx在线观看| 成人国产精品一级毛片视频| 国产欧美一区二区精品忘忧草| 亚洲欧美日韩网| 国产成人综合美国十次| 伊人久久大香| 国产成人亚洲综合a∨猫咪| 精品国偷自产国产一区| 狠狠艹夜夜干| 亚洲福利影视| 国产精品一二一区| 精品成人一区二区| 天天干人人干| 精品中文字幕一区二区三区四区 | 国产图片综合| 中国色在线日|韩| 日韩不卡一区二区三区| 欧美精品黑人性xxxx| 尤物视频网站| 国产福利一区二区三区在线播放| 久久福利视频一区二区| 欧美一区二区视频网站| 美女被艹网站| 成人久久精品| av网站一区二区三区| 国产性猛交xxxx免费看久久| 最猛黑人系列在线播放| 久久93精品国产91久久综合| 国产精品欧美极品| 欧美另类极品videosbest最新版本 | 99精品免费视频| 欧美在线免费视屏| 一级美女在线| 四虎影视精品永久在线观看| 国产91综合一区在线观看| 日韩电影免费观看中文字幕| 亚州黄色一级| heyzo久久| 亚洲一区二区三区激情| 亚洲狠狠婷婷综合久久蜜桃| 人人草在线视频| 久久国产精品99久久久久久老狼 | 天天综合av| 国产综合色精品一区二区三区| 亚洲第一区中文字幕| 欧美成人精品一区二区| 亚洲伦伦在线| 91高清在线| 一区二区三区高清在线观看| 91女厕偷拍女厕偷拍高清| 在线观看国产精品日韩av| 在线91av| 91成人影院| 色天天综合久久久久综合片| 91香蕉视频在线观看| 亚洲综合资源| 久久久久国产一区二区三区四区| 欧美超级免费视 在线| 国内精品久久久久国产| 亚洲女同在线| 精品精品国产高清a毛片牛牛 | 国产欧美三级电影| 国产精品国产精品国产专区不片| 亚洲天堂一区| 樱桃视频成人在线观看| 成人动漫av在线| 欧美插天视频在线播放| 任你弄在线视频免费观看| 日本免费在线视频不卡一不卡二| 亚洲国产精品女人久久久| 中文字幕一区免费| 欧美另类亚洲| 欧美一区二区成人6969| 一级毛片免费看| 欧美三级视频| 欧美一区二区国产| 欧美18一12sex性处hd| 欧美国产专区| 日韩午夜在线观看| 人成福利视频在线观看| 欧美日韩一区自拍 | 日韩精品有码在线观看| 黄色小视频在线免费观看| 亚洲一区欧美二区| 亚洲精品电影网站| 牛牛热在线视频| 麻豆九一精品爱看视频在线观看免费| 精品福利一区二区三区免费视频| 在线看小视频| 亚洲经典自拍| 日韩国产中文字幕| 69av亚洲| 精品亚洲aⅴ乱码一区二区三区| 中文字幕久久精品| 在线三级中文| 福利电影一区二区三区| 色综合久综合久久综合久鬼88| av资源在线| 99精品久久免费看蜜臀剧情介绍| 午夜精品一区二区三区在线视 | 亚洲精品免费看| 国产精品久久在线| 欧美美女啪啪| 亚洲va天堂va国产va久| 青青色青青操| 国产精品亚洲片在线播放| 欧美视频在线视频| 国产精选在线视频拍拍拍| 成人激情自拍| 偷窥少妇高潮呻吟av久久免费| av免费在线播放| 久久裸体网站| 欧美日韩电影在线| 免费白浆视频| 国产农村妇女精品一区二区| 日韩电影免费观看在线观看| 麻豆传媒免费在线观看| 激情六月婷婷综合| 久久福利网址导航| 欧美性理论片在线观看片免费| 国产丝袜欧美中文另类| 最近2018年中文字幕在线| 五月天色综合| 一区二区三区加勒比av| sese视频| 亚洲精品久久| 亚洲精品一区二区三区影院| 成人精品一区二区三区校园激情| 青青国产91久久久久久| 欧美成人剧情片在线观看| 樱桃视频成人在线观看| 亚洲欧洲三级电影| 性生活视频网址| 欧美理论视频| 精品欧美一区二区三区精品久久| 精品亚洲综合| 国产在线播放一区二区三区| 色综合久综合久久综合久鬼88| 天然素人一区二区视频| 一区二区三区加勒比av| 天天看天天干| 欧美极品一区二区三区| 亚洲成色www8888| 婷婷激情在线| 久久亚区不卡日本| 九色蝌蚪国产| 欧美综合一区| 精品奇米国产一区二区三区| 午夜不卡视频| 91在线精品一区二区| 欧美人与动性xxxxx杂性| 欧美日韩在线二区| 欧美白人最猛性xxxxx69交| 国内精品久久久久久野外| 久久久亚洲午夜电影| 精品国产免费观看一区| 特黄特色欧美大片| 欧美日本免费一区二区三区| 青青青手机在线视频观看| 国模大尺度一区二区三区| 四虎激情影院| 亚洲都市激情| 精品国产免费一区二区三区四区| 在线免费av导航| 国产精品久久久久久一区二区三区| 碰碰视频免费| 午夜天堂精品久久久久| 日韩在线视频观看| 一区在线影院| 黑丝美女久久久| 一级黄色在线| 国产成人免费av在线| 国产免费自拍视频| 99久久久久| 在线精品国产成人综合| 国产成人精品亚洲日本在线观看| 欧美视频专区一二在线观看| 欧美91精品久久久久国产性生爱| 国产成人啪免费观看软件| 97超级碰碰碰久久久| 女仆av观看一区| 精品免费国产二区三区| 俺来也官网欧美久久精品| 国产精品免费观看视频| 男人艹女人网站| 日韩不卡一区二区| 色精品一区二区三区| 99久久国产综合精品成人影院| 色小说视频一区| www久久久|