Elon Musk doesn’t apologize often. So when he does that, people pay attention.
In a Thursday post on He also apologized to candidates the company passed on in error and said he and xAI head of talent Baris Akis were combing through old interview transcripts to reconnect with people who deserved opportunities.
The admission is shocking for a company founded in 2023 with the promise of unlocking the mysteries of the universe. This is even more remarkable given the timing.
Just six weeks ago, SpaceX acquired xAI, valuing the combined entity at $1.25 trillion. Previously, Tesla disclosed a US$2 billion investment in xAI’s Series E round of financing in the fourth quarter of 2025. Now, Musk is telling the world that what he just sold his investors is broken.
More AI stocks:
Tesla shareholders have sued Musk for breach of fiduciary duty, saying he diverted artificial intelligence talent and resources away from Tesla to benefit his private business. This admission adds a new dimension to these legal challenges.
Meanwhile, SpaceX is gearing up for a record-setting IPO later this year. A faltering AI unit isn’t the story Musk needs investors to read right now.
Of the 12 people who co-founded xAI with Musk in 2023, only two remain. Manuel Kroos and Ross Nordin were the last ones standing.
In early 2026, departures increased dramatically. Here are the people who left.
-
Jimiba: He was one of xAI’s most prominent artificial intelligence researchers, and he reportedly left the company in February due to model performance issues.
-
Tony Wu: He left the same week as Tomoe with no public explanation given.
-
Toby Pollan: Responsible for the ambitious “Macrohard” coding project, he left after just 16 days on the job.
-
Zhang Guodong: Leads the Imagine team at xAI. His departure was confirmed this week. Reuters reported that Musk accused him of coding flaws in the product.
-
Dai Zihang: Investigate Grok’s coding capabilities. left earlier this week, Reuters reported.
Exiting is not just a people issue. Insiders describe a combination of burnout, Musk’s management style and an organizational structure that was never built to sustain the kind of aggressive AI development the company promised.
