A Chinese dissident based in the UK has faced harassment via a series of deepfake posts that portray her inaccurately as a drug addict. This incident has raised significant concerns and discussions regarding how social media platforms, notably one owned by Elon Musk, manage AI-generated content, particularly when it involves political figures critical of oppressive regimes.
Apple Peiqing Ni is the founder of the China Dissent Network, dedicated to backing Chinese activists residing in the UK. She has actively participated in UK parliamentary events focused on the 1989 Tiananmen Square crackdown. Following her posts on these events, she quickly became a target of a coordinated campaign of harassment utilizing manipulated AI-generated images.
The campaign against Ni featured at least twelve deepfake images, all crafted to undermine her credibility and present her in a disturbing light, falsely depicting her as involved in drug use and promiscuity.
After reporting the matter to the police, she was advised to reach out to the social media platform for help. In response, the platform concluded that the content did not breach their community standards.
This series of deepfake posts began appearing just after her commentary on the Tiananmen incident, indicating a potential link to accounts seemingly aligned with Chinese state interests. Researchers studying transnational online repression have noted a pattern whereby vocal dissidents face harassment through such tactics.
The platform does have policies against sharing non-consensual and manipulated images. However, this situation raises questions about enforcement and the effectiveness of their content moderation algorithms. It suggests either a failure in recognizing blatantly fabricated material or a significant narrowing of the definitions used in determining violations.
As regulatory pressure mounts in Europe, with new laws like the UK’s Online Safety Act urging platforms to protect users from harmful content, cases like Ni’s highlight potential failures in compliance. European authorities are scrutinizing how platforms manage deepfake content and may impose severe fines if they find systemic neglect in removing non-consensual imagery.
Ultimately, for Ni, the personal ramifications are severe. As a young activist who spoke out about a historical atrocity, the reaction she faced involved fabricated abusive imagery, and the platform that could have offered protection deemed that behavior acceptable within its guidelines.