AI Risk Wire

Cases › produced harmful content › LR-2026-0769

Charges over AI child abuse material allegedly created using Grok

Preliminary Why: Publisher not yet named · subject to change

Status: no action reported

Severity moderate
Reported 27 Oct 2026
Developer and model xAI · Grok

No source states: loss or records, agent tool, could do, deployer, industry.

Severity moderate: tier: data of others (covert recordings of women and teenage girls) touched with a bounded scope stated as more than 150 recordings found on the seized devices.

A man from Pennsylvania, Robert Mayberry, has been charged with child sexual abuse material and invasion of privacy offences. Police allege that in one case he used Grok, an AI service developed by xAI, to create abusive material from a photograph of a teenage girl taken in a public park. Investigators allege that some material found on his devices was AI-generated, and they also reported more than 150 covert recordings of women and teenage girls. The investigation began on 18 June 2026 after a woman reported concerning images on his telephone, and police seized his devices under a search warrant. Mayberry was arrested on 21 October 2026 and posted bail, and a preliminary hearing is scheduled for 30 October 2026. The allegations have not been proven in court, and the source does not establish which Grok version, settings or safeguards were involved.

What happened

1 · What it was doingPolice allege a man photographed a teenage girl in a park and used Grok to create abusive material from the photograph.
2 · How it was caughtA woman reported concerning images on his telephone and police began an investigation.

What was done

No source states an action or fix.

Risks

Harmful content and jailbreaks
Primary
84 incidents
Rising: 6 vs 3.8 a week

Sorted by rule from the case's fields; not yet read from the sources.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Model

Safety training and refusal · Content safety classifiers · Abuse monitoring and account enforcement · Outside red-teaming

Application

Output filtering and redaction

Similar incidents

All cases of this kind ›

Evidence

No source states: what went wrong, what it reached or did, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off, what was done.

Show the full account: 4 items from the primary sources

What happened, in order

  1. 2026-06-18 Police began investigating after a woman reported concerning images on Mayberry’s telephone.
  2. 2026-10-21 Mayberry was arrested and subsequently posted bail.
  3. 2026-10-30 A preliminary hearing is scheduled.

Not dated in the sources:

  • Police obtained a search warrant and seized electronic devices.
Show every field, with its source and grade

The AI system

Model
Grok
C one secondary source states it
Developer
xAI
C one secondary source states it

What happened

Earliest source we hold
2026-10-27
D inferred by us from the described behaviour, reason shown
What went wrong
produced harmful content
C one secondary source states it

Impact

Severity
moderate
D inferred by us from the described behaviour, reason shown
Why this severity
tier: data of others (covert recordings of women and teenage girls) touched with a bounded scope stated as more than 150 recordings found on the seized devices.
D inferred by us from the described behaviour, reason shown

Notes on gaps

All facts are allegations attributed to police and NBC10 reporting; no court finding exists yet. No date of occurrence or date of disclosure is stated in the article text; the article's publication date of 2026-10-27 was not used as a disclosure date. The 2024 Pennsylvania law and FBI warning are background and were not treated as facts about this incident. Durations, money, and record counts are not stated and were left out.

LR-2026-0769 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON