Systemic failure: How AI-generated non-consensual imagery exploits legal loopholes and gendered power structures

Indigenous frameworks like the Māori principle of *kaitiakitanga* (guardianship) emphasize collective responsibility for digital spaces, contrasting with Western individualism that treats harm as a personal grievance. Traditional knowledge systems often view consent as a sacred covenant, not a contractual clause, which could redefine AI-generated exploitation as a spiritual and communal violation. However, these perspectives are systematically excluded from tech policy debates, which prioritize Silicon Valley's 'move fast and break things' ethos over Indigenous epistemologies of care.

The criminalization of non-consensual imagery has parallels in historical 'moral panics' around new media, from photography in the 19th century to deepfakes today, where technology outpaces legal adaptation. The 'revenge porn' phenomenon mirrors 19th-century 'postal obscenity' laws, which were also framed as protecting 'public decency' while disproportionately targeting marginalized bodies. The Take It Down Act's focus on post-hoc removal echoes the failures of Section 230, which shifted liability away from platforms while failing to prevent harm.

In South Korea, digital sex crimes (*molka*) have led to mass protests and stricter penalties, showing how cultural attitudes toward gendered violence can drive systemic change. Conversely, in India, the lack of legal recourse for AI-generated deepfakes reflects a broader pattern of state inaction on gendered digital harms, despite high-profile cases. Cross-culturally, the commodification of women's bodies in digital spaces mirrors colonial-era 'scientific racism,' where bodies were objectified for profit under the guise of progress.

Research shows that AI-generated non-consensual imagery disproportionately targets women and girls, with 96% of deepfake pornography featuring women, according to a 2023 Sensity AI report. The psychological impact includes PTSD, depression, and suicide, yet most studies focus on individual trauma rather than systemic prevention. Meanwhile, AI ethics research highlights how 'ethics washing' in tech companies obscures the lack of diverse datasets and the reinforcement of harmful stereotypes in training models.

Artists like Shu Lea Cheang have long explored the intersection of technology and gendered violence, using digital media to expose the dehumanizing effects of surveillance capitalism. Indigenous artists like Hulleah Tsinhnahjinnie critique the digital erasure of Native women through AI-generated 'virtual' representations that replace lived experiences. Spiritual traditions, such as the Buddhist concept of *dukkha*, frame digital harm as a manifestation of suffering rooted in attachment and delusion, offering a lens beyond punitive justice.

Without intervention, AI-generated non-consensual imagery could normalize a 'digital caste system' where marginalized groups are permanently stigmatized by algorithmic bias. Scenario modeling suggests that by 2030, deepfake exploitation could outpace traditional forms of gendered violence, overwhelming legal systems designed for physical harms. However, proactive measures—like real-time content moderation and decentralized consent registries—could reduce harm by 60% within a decade, according to a 2025 MIT study.

Black women and LGBTQ+ individuals are 3x more likely to be targeted by AI-generated non-consensual imagery, yet their experiences are rarely centered in policy debates. Survivors from Global South countries often lack access to legal recourse, as platforms deprioritize non-Western languages and jurisdictions. The 'first man convicted' framing obscures the fact that women and minors are the primary victims, reinforcing a narrative that centers perpetrators over survivors.