Elon Musk's xAI: Controversial Training on Claude's Data
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
The company xAI, founded by Elon Musk, reportedly used the results of Claude, a model developed by Anthropic, to train its own coding models. According to The Information, this practice lasted several months before Anthropic officially cut off access last January.
Despite this cutoff, xAI allegedly continued to use Claude's data through personal accounts and the intermediary service Blackbox AI. Elon Musk has already acknowledged in court that xAI partially used models from OpenAI to develop Grok, emphasizing that this is common in the industry.
The company is facing notable internal difficulties. The pre-training team has been reduced to fewer than five members, and four Grok code leads have recently left the company, along with several co-founders. An internal incident also led to the accidental deletion of crucial training data, resulting in a loss of two to three weeks of work.
Moreover, Musk, who had planned a significant breakthrough for fall 2025, now finds himself renting computing power from SpaceX to Anthropic and Google, instead of using it for his own models. This situation is being presented as temporary.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.