XAI sued for allegedly training Grok models with child sexual abuse material

XAI, the artificial‑intelligence research firm behind the language model Grok, has come under scrutiny after a whistleblower alleged that the model was trained on a dataset containing both real and AI‑generated child‑pornographic material. According to the complaint, the dataset was compiled without proper safeguards or verification protocols, raising concerns about the ethical and legal compliance of XAI’s training practices.

The allegations were first reported by a former XAI employee who claimed that internal data‑curation teams had accessed and incorporated illicit content into Grok’s training corpus. XAI has denied the accusations, stating that all data used for training were sourced from publicly available and licensed datasets and that it employs strict content‑filtering mechanisms. The company has also announced an independent audit to review its data‑handling procedures and to ensure compliance with child‑protection laws.

Regulators and child‑protection advocates are monitoring the situation closely, noting that the use of such material could violate multiple jurisdictions’ statutes on child exploitation and digital content. The outcome of the audit and any subsequent investigations will determine whether XAI faces legal penalties or must overhaul its data‑collection protocols to prevent future breaches.