French authorities probe X over algorithm interference and harmful content

What happened

French prosecutors summoned Elon Musk and former X CEO Linda Yaccarino for a voluntary hearing on April 20, 2026, as part of a cybercrime probe. Both Musk and Yaccarino failed to appear.

The investigation was launched in January 2025 following a report by French politician Éric Bothorel raising concerns over X's algorithms and their alleged influence on automated data processing systems. In February 2026, French police conducted a raid on X's Paris headquarters in partnership with Europol.

The probe was expanded in February 2026 to include an investigation into X's Grok chatbot's creation and dissemination of Holocaust denial material, which is a crime in France. The investigation also covers millions of sexualized deepfake images that proliferated on X from late December 2025 to early January 2026 following the introduction of a new image editing function.

Reported impact

French prosecutors suspect that Musk encouraged the creation of deepfakes potentially to inflate the value of X. The dissemination of Holocaust denial content and sexually explicit deepfakes raised concerns about X's compliance with French law and content moderation practices.

Index analysis

The investigation focused on whether X's algorithms were used to interfere in French politics and whether the platform failed to moderate harmful content. Prosecutors characterized the inquiry as a "constructive approach" aimed at ensuring X complies with French law.

Governance controls

Following backlash over the sexually explicit deepfake material, X introduced restrictions preventing Grok from editing images of real people to show them undressed or in revealing clothes. X dismissed the investigation as politically motivated.

Classification

Organizations

X (subject), OpenAI (subject), Paris Prosecutor's Office (regulator), Elon Musk (subject), Linda Yaccarino (subject)

Category

Misinformation and synthetic media

Industry

Media, social platforms and marketing

Country

France

Harm status

Actual harm

Intent

Unknown

Timing

Post-deployment

Harm types

Public interest, Human or fundamental rights, Reputational

Harmed

General public, Government

Failure modes

Deepfake image, Content moderation failure

Consequences

Regulatory investigation

AI systems

Grok

Regulations

GDPR (investigation)

Sources and evidence

Related incidents