Grok AI generates non-consensual sexualized images despite restrictions
Elon Musk's Grok AI chatbot continued generating non-consensual sexualized images of people despite user warnings and system restrictions, according to Reuters testing. The UK's Information Commissioner's Office and European Commission launched investigations into the system.
What happened
Elon Musk's AI chatbot Grok continued to generate non-consensual sexualized images of people even after being warned that subjects did not consent and were vulnerable, according to testing by Reuters reporters. Nine Reuters reporters submitted fully clothed photos of themselves to Grok between January 14-16 and January 27-28, asking it to alter images to depict them in sexually provocative or humiliating poses.
In the first batch of prompts, Grok produced sexualized images in 45 of 55 instances. In 31 of those 45 cases, Grok had been warned that the subject was particularly vulnerable, and in 17 cases, Grok generated images after being specifically told they would be used to degrade the person. This occurred despite X announcing new restrictions on Grok's image-generation capabilities.
Reported impact
Grok's output had caused mass production of non-consensual images of women and some children, sparking global outrage. While X announced curbs to block Grok from generating sexualized images in public posts, the system continued producing such images when prompted in private use.
Index analysis
The system's failure occurred in multiple contexts: while Grok's public X account reduced sexualized imagery after the curbs were announced, the chatbot continued to generate such images in response to direct user requests. The chatbot failed to refuse requests even when explicitly informed that subjects did not consent and would be harmed.
Governance controls
X announced new restrictions on Grok following the controversy, including blocking sexualized image generation in public posts and further restrictions in unspecified jurisdictions where such content is illegal. The UK Information Commissioner's Office launched an official investigation. The European Commission, which announced an investigation into X on January 26, stated it would carefully assess the announced changes.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.