AI Companies Must Face Accountability for Harmful Technology

Grok has caused significant damage recently. Earlier in the year, it produced millions of deepfakes on an unprecedented scale. This situation led to legal action and sparked a national discussion on what AI companies owe to people harmed by their products. More recently, Grok exposed the identities of sex workers. According to 404 Media, Grok disclosed the legal name of pornographic actress Siri Dahl without her consent or warning. This resulted in harassment and identity impersonation. Her privacy disappeared almost instantly.

Online porn performers understand privacy deeply due to the stigma, discriminatory policies, and potential violence they face. For many sex workers, privacy is essential, not optional. They use pseudonyms and often hide their faces in images. They invest in keeping their professional and legal identities separate. Exposure can lead to loss of housing, custody issues, or physical harm. An AI chatbot undoing this privacy in seconds represents a crisis.

Dahl noted that her legal name only became accessible after Grok revealed it. The chatbot did not merely reveal existing information; it caused a new exposure. While sex workers experience these threats severely, everyone is vulnerable. AI chatbots can amplify harm. Futurism reported incidents where chatbots inspired violence, sexual abuse, and stalking. Unlike search engines, chatbots don’t just provide results. They engage and reinforce. They can validate obsessions and paranoia. When combined with access to real names, addresses, and family details, chatbots become tools for aiding stalkers.

We must control what we can. Data broker sites buy, sell, and publish personal data. Opting out is possible but tedious. Many use subscription services to monitor and request removals. If you haven’t considered this, you should. Requests might take 90 to 120 days. Delay helps others find you.

Personal vigilance is limited. AI chatbots can unravel digital precautions, connecting seemingly unrelated data. Companies responsible for harmful tools must face legal consequences. Liability laws apply to such cases. When products cause harm, novelty or unintended outcomes aren’t excuses. Advocates and lawyers are starting to challenge AI tools’ accountability. Legal actions have already affected the tech sector. Litigation against Omegle, which enabled exploitation, led to its shutdown in 2023.

The philosophy is simple: If technology foreseeably causes harm, the responsible companies must answer for it. Tech firms often argue they just host others’ information. However, when AI actively exposes personal data, this defense weakens. Sex workers highlight digital safety risks. The question remains if society and courts will act before these tools harm others. Defective AI tools must hold their creators to account.

Norma Buster is the chief of staff at C.A. Goldberg PLLC, a law firm focused on justice for victims of harmful tech. She hosts the podcast “Oral Arguments,” engaging with leaders in sex, technology, and victims’ rights intersections.