When technology fails women: Online abuse and Nigeria's digital weak points

• Generative AI tools like xAI's Grok are intensifying existing patterns of online gender-based violence (OGBV) by enabling the rapid creation and weaponization of non-consensual sexualized imagery.
• In Nigeria, 58% of online abuse cases target women, and only 24% of citizens view platforms like X as responsive to harm reports.
• Projections suggest up to 30 million Nigerian women and girls could be directly targeted by AI-facilitated abuse annually by 2030.

Structural Vulnerabilities and AI

• Nigeria's regulatory landscape for AI is fragmented across multiple agencies, resulting in limited accountability and enforcement for platform-based harm.
• X's monetization model rewards viral engagement, which can inadvertently incentivize perpetrators to amplify AI-generated harassment for profit.
• Human-driven abuse existed long before AI; however, generative tools have systematized and scaled these harms, lowering the barrier for creators of exploitative content.

The Role of Design in Safety

• UX design choices regarding data retention, privacy, and moderation directly impact user safety and the proliferation of passive-aggressive or abusive content.
• The "Gendered Privacy Evaluation Framework," developed by Superbloom and TechSocietal, provides a mechanism for companies to assess if their AI systems reinforce or reduce gendered harm.
• Effective mitigation requires moving beyond surface-level safety features toward deep-system changes, including meaningful consent, data minimization, and structured stakeholder engagement with women's rights groups.