人工智能技术正日益加剧针对女性和女童的性暴力与形象滥用问题。据Security Hero 2023年的研究显示,2019年至2023年间,在线深度伪造视频数量激增550%,其中98%的内容属于色情性质,且99%以女性为目标1。这一现象在澳大利亚同样严峻,ANROWS 2022年的研究指出,超过半数的受访澳大利亚成年人曾遭受技术促成的虐待,其中18至24岁年轻群体的遭遇比例高达71.5%,四分之一的受访者经历过涉及性或形象的滥用行为1。
除了受害者人数众多,AI行业内部存在的性别失衡也被视为问题根源之一。联合国妇女署在2025年11月发布的报告中揭示,全球仅有30%的AI专业人员为女性;此外,在评估的138个国家中,只有24个国家的国家AI战略提及了性别议题,而包含实质性应对措施的国家更是仅占18个1。面对此类危机,澳大利亚监管机构已采取行动,网络安全专员Julie Inman Grant警告称,“nudify”类应用在澳大利亚每月的下载量约为20万次,其办公室已对8家提供此类服务的机构实施了执法行动1。
为了从法律层面遏制危害,澳大利亚政府已于上月向议会提交《数字注意义务法案》(Digital Duty of Care legislation)草案1。该法案旨在要求在线服务商防止危害并强化监管权力1。
The proliferation of deepfake technology has fueled a sharp increase in online sexual abuse, with the vast majority of such content targeting women. Research conducted by Security Hero indicates that between 2019 and 2023, the number of online deepfake videos surged by 550%, with 98% classified as pornographic and 99% featuring women as victims 1. This trend is compounded by broader issues of tech-facilitated harm; a 2022 study by ANROWS revealed that over half of surveyed Australian adults have experienced such abuse, a figure rising to 71.5% among those aged 18 to 24, while one-quarter reported experiencing sexual or image-based abuse specifically 1.
In response to these escalating risks, Australian authorities are taking concrete regulatory steps. Julie Inman Grant, the Australian eSafety Commissioner, warned that "nudify" applications are downloaded approximately 200,000 times per month within the country, prompting her office to initiate enforcement actions against eight such services 1. Furthermore, the Australian government introduced the Digital Duty of Care legislation to parliament last month, aiming to mandate that online service providers prevent harms and grant regulators stronger powers to address these violations 1.
Underlying this crisis are significant structural imbalances within the artificial intelligence sector itself. A November 2025 report from UN Women highlights that women constitute only 30% of AI professionals globally 1. The same report noted that out of 138 countries assessed, merely 24 mentioned gender considerations in their national AI strategies, and only 18 included substantive measures to address these disparities 1.
评论
还没有评论,欢迎留下第一条。