Grok, developed by Elon Musk's xAI, introduced a 'spicy mode' option that enables the chatbot to generate responses with fewer content restrictions. According to BBC News and The Guardian, this setting was advertised to provide more unfiltered and sometimes provocative answers compared to standard output.
After regulatory investigations and media scrutiny over Grok's generation of sexualized deepfake images, the role of 'spicy mode' came under particular attention. While 'spicy mode' was not the sole reason for problematic content, its reduced filtering made it more likely that users could prompt Grok to produce controversial or inappropriate material.
Following these incidents, xAI revised its policies and temporarily disabled Grok's image generator, including the 'spicy mode' feature, as confirmed by coverage in PBS and FRANCE 24. This move was part of broader efforts to address regulatory concerns and improve content moderation on the platform.