Skip to content
Submit incident
AII-1001-04343
Under investigationA regulator, court or police has opened a formal inquiry and reached no finding yet.

French authorities investigate Grok AI for generating non-consensual deepfake images

French prosecutors launched an investigation into Grok after parliamentarians complained that the AI system generated approximately 3 million non-consensual sexual deepfake images between late December 2025 and early January 2026, including 23,000 depicting minors. Despite xAI's stated restrictions starting 9 January 2026, tests showed the system continued generating explicit imagery of male subjects. Elon Musk and X executives were summoned for questioning in April 2026.

January 2026 · xAI · Harmful and abusive content
January 2026
AII-1001-043434 sources ↓
Quick facts
Category
Harmful and abusive content
Country
France
Source
DAGOSPIA
Date
January 2026
Drafted by
Raidu Incident Detection

What happened

The Paris prosecutor's office opened an investigation into Grok after French parliamentarians Arthur Delaporte and Eric Bothorel filed a complaint in January 2026 about thousands of non-consensual sexual deepfake images generated by the AI system.

A Center for Countering Digital Hate analysis found that between 29 December 2025 and 9 January 2026, approximately 3 million sexually explicit images of real people were produced or modified on Grok, including 23,000 depicting minors.

On 9 January 2026, xAI limited the ability to generate nude images from real photos to X platform subscribers; five days later this restriction was extended to all users following international government and institutional protests.

Journalist Robert Hart reported that despite xAI's stated restrictions, Grok continued to generate sexually explicit images of male subjects, including fetish imagery and nude companions.

Reported impact

The creation of non-consensual deepfake images depicting real individuals in explicit sexual scenes raised concerns about privacy violations and psychological harm to victims, particularly women and minors.

Index analysis

The incident highlights failures in content moderation and AI safeguards, as restrictions implemented by xAI proved incomplete and could be circumvented using alternative formulations or targeting different demographics.

Governance controls

The Paris prosecutor announced a constructive approach aimed at ensuring X platform's compliance with French law. Elon Musk and Linda Yaccarino were summoned for questioning scheduled 20–24 April 2026, with X platform employees also to be heard as witnesses.

The European Union initiated a separate investigation into Grok's deepfake generation capabilities in late January 2026.

Classification

Category
Harmful and abusive content
Industry
Media, social platforms and marketing
Country
France
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Human or fundamental rights, Reputational, Psychological
Harmed
General public, Children, Women
Failure modes
Deepfake image, Content moderation failure
Consequences
Regulatory investigation
AI systems
Grok
Regulations
EU AI Act (investigation)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

DAGOSPIA
hard grok! - il programma di intelligenza artificiale di 'x', il social di elon musk, continua a cre
3 February 2026
Forbes Italia
La Procura di Parigi perquisisce l'ufficio francese di X. Musk convocato per un "interrogatorio volontario
3 February 2026
Il Fatto Quotidiano
Elon Musk convocato dalla procura di Parigi: indagine sui deepfake di Grok - Il Fatto Quotidiano
3 February 2026
Tom's Hardware
Immagini AI, minori e responsabilità penali: il caso Grok davanti all'Europa
6 February 2026
Oct 2026AI deepfakes and harassment campaigns target Nigerian women, report warnsReportedHarmful and abusive contentPremium Times NigeriaLeadershipSep 2026UK ASA bans AI app ads for objectification and explicit contentConfirmedHarmful and abusive contentBBCAug 2026Woman Joins Lawsuit Against xAI Over Grok's Role in Creating CSAM DeepfakesAllegedHarmful and abusive contentخبرآنلاینخبرگزاری مهر | اخبار ایران و جهان | Mehr News Agencyايتنا - سایت خبری تحلیلی فناوری اطلاعات و ارتباطاتAug 2026Wyoming woman alleges Grok used to generate thousands of explicit deepfake images from her childhood photoAllegedHarmful and abusive contentwashingtonpost.comAug 2026Woman sues xAI over Grok use to generate explicit images from childhood photoAllegedHarmful and abusive contentTechCrunchWashington PostTimesNow