xAI sues Grok user over sexualized deepfakes as child-exploitation class action expands

xAI, Elon Musk's AI company, filed a lawsuit against a Grok user identified as Harwood — arrested in February — over sexualized 'deepfakes' generated using the tool, according to Reuters. Simultaneously, a separate child-exploitation class action against xAI expanded, adding new plaintiffs and naming Stability AI as a co-defendant, deepening the legal jeopardy around image-generation misuse.

The dual legal fronts capture the bind image-generation tools face: xAI is positioning itself as the victim/enforcer by suing a user who allegedly misused Grok, while simultaneously being sued as the enabler of harmful content. Critics argue that lax guardrails on Grok's image generation made such misuse foreseeable, and the class action naming Stability AI signals plaintiffs are pursuing the model providers, not just end users.

The legal heat compounds a rough stretch for Grok's reputation even as the model gains distribution — Grok 4.3 landed on Amazon Bedrock the very next day, and FourWeekMBA analyzed Grok 4.5 as xAI's first coding-and-agent-focused model. The tension between rapid capability/distribution gains and mounting safety litigation is the story to watch: enterprises evaluating Grok on Bedrock will weigh the model's performance against the reputational and legal cloud around its consumer image tools. For the broader industry, the Stability AI co-defendant move could set precedent on how far liability extends up the model-provider chain.