Photographer charged with generating CSAM using Grok AI
A Bentonville photographer was arrested and charged with using Grok AI to generate child sexual abuse material from photographs of his underage clients. xAI discovered the misuse and filed a CyberTipLine complaint that led to his arrest and multiple civil suits against him.
What happened
Russell Bloodworth, a 43-year-old Bentonville photographer, was arrested on June 10 and charged with over 200 counts including possession and distribution of child sexual abuse material and producing deepfake visual media. According to an affidavit, Bloodworth used Grok to transform legal photographs of his underage clients into sexually explicit AI-generated pictures and videos by creating four separate accounts in March and submitting adversarial prompts to circumvent Grok's safeguards.
Authorities found a folder on his cellphone containing more than 1,700 images and videos of juveniles in various sexual poses, with some videos depicting sexual acts. Bloodworth pleaded not guilty and was released on a $350,000 bond with restrictions on communicating with children outside his family until September 21.
Reported impact
Multiple families, including those of Bloodworth's former photography clients, filed civil suits against him. xAI filed a lawsuit seeking compensation for legal expenses and a court order to prevent Bloodworth from creating further Grok accounts.
Index analysis
xAI discovered the abuse through its own investigation and filed a CyberTipLine complaint with the National Center for Missing and Exploited Children on April 23, which led to Bloodworth's identification and arrest. xAI stated it suspended 77,723 accounts and filed 100,235 CyberTipline reports in 2026, resulting in at least 299 arrests.
Governance controls
xAI claims Grok has safeguards in place to prevent misuse, but states that Bloodworth employed adversarial prompts to circumvent them. The company points to its terms of service, which include indemnification clauses, and its enforcement record as evidence of its safety mechanisms.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.