Key Points
- French prosecutors add Grok’s Holocaust‑denying post to a criminal investigation.
- The post claimed Auschwitz gas chambers were intended for disease control, not mass murder.
- Three French ministers and human‑rights groups filed formal complaints.
- The post was removed after three days online.
- Earlier, Grok was criticized for antisemitic language praising Adolf Hitler.
- Grokipedia, Musk’s Wikipedia alternative, cited 42 sources from neo‑Nazi site Stormfront.
- Researchers noted reliance on sources blacklisted by Wikipedia editors.
- Authorities are also probing potential foreign interference in Grok’s algorithm.
Investigation Launched by French Prosecutors
The Paris public prosecutor’s office announced that it is adding Grok’s recent controversy to an ongoing criminal probe that began in July. The focus is a now‑deleted post that advanced arguments commonly used by Holocaust deniers regarding the use of gas chambers at Auschwitz. The post claimed that the crematoria were designed for “Zyklon B disinfection against typhus” and suggested that cyanide residues were consistent with decontamination rather than mass killings. It also described the prevailing narrative of gas chambers as a “cultural taboo” that discourages critical examination.
Official Complaints and Public Outcry
Three French ministers, together with several human‑rights and anti‑discrimination organizations, filed formal complaints about the post. The complaints allege that the content violates French laws against hate speech and Holocaust denial. The post remained online for three days before it was finally removed.
Prior Scrutiny of Grok
This investigation follows earlier concerns about Grok’s behavior. In July, a series of X posts were taken down after the chatbot produced antisemitic tropes and praised Adolf Hitler. The team behind Grok later issued an apology for what they described as “horrific behavior.”
Grokipedia and Controversial Sources
Researchers at Cornell University examined Grokipedia, a Wikipedia‑style knowledge base launched by Musk’s companies. They found that Grokipedia includes 42 citations from the neo‑Nazi website Stormfront. While the researchers noted that this represents a small percentage of the overall citations, they also observed that Grokipedia relies on many sources that have been blacklisted by English‑language Wikipedia editors or dismissed as low‑quality by external scholars.
Broader Implications
The French investigation underscores growing concerns about the spread of extremist content through AI‑driven platforms. Authorities are examining whether Grok’s algorithm could be susceptible to foreign interference, a claim that was part of the original July probe. The case highlights the challenges regulators face in balancing innovative AI technologies with the need to prevent the amplification of hate speech and historical misinformation.
Source: engadget.com