The question of legal responsibility when Artificial Intelligence (AI) causes harm took center stage on July 7, 2026, during the second day of the inaugural UN summit on AI governance. Leading experts presented mounting evidence of human rights violations linked to AI technology.

Sonia Livingstone of the Independent International Scientific Panel on AI reported that across 11 Global South countries, up to one child per classroom indicated AI was used to create sexually explicit deepfakes of them. She also noted that reports of AI-generated child sexual abuse material to the US CyberTip line, NCMEC, are rising exponentially.

Livingstone emphasized that evidence of AI-related violations against individuals and vulnerable groups is "much more compelling" than the benefits, citing the technology's use to "create and amplify persuasion and deceit, to spread disinformation, distrust," and the exploitation of personal data by self-learning systems.

Morocco’s Amal El Fallah Seghrouchni stressed the need to "open up the black boxes of AI," explaining that large language models operate continuously using around 180 billion parameters, yet few lawyers or sociologists fully understand their workings.

A survey commissioned by UN Women revealed that one in four women human rights defenders, activists, and journalists experienced AI-assisted online violence, with six percent reporting victimization by deepfakes or digitally manipulated imagery. Sima Bahous, Executive Director at UN Women, noted much of this abuse remains undocumented and unreported.

The Independent International Scientific Panel on AI also reported that 88 percent of leading AI researchers are male, highlighting gender imbalance in the field.

Volker Türk, UN High Commissioner for Human Rights, advocated for embedding equality, accountability, and human oversight into AI's design, development, and deployment to ensure the technology serves humanity.

The UN calls for inclusive global participation in AI governance to address both the harms and benefits of AI technology.

Sources