Elon Musk's artificial intelligence startup, xAI, is reportedly in talks with investors to raise a fresh round of funding, which could value the company at approximately $40 billion. According to sources familiar with the discussions, xAI aims to secure several billion dollars in this funding round, though the negotiations are still in early stages, making terms subject to change.
Earlier, in May, xAI completed a series B funding round, raising $6 billion and achieving a post-money valuation of $24 billion. This funding was led by prominent investors such as Andreessen Horowitz and Sequoia Capital.
Musk founded xAI in July 2023 as a competitor to OpenAI's ChatGPT. In a move towards open-source innovation, xAI introduced its ChatGPT alternative, "Grok," in March, giving the public free access to the underlying technology. This aligns xAI with other tech companies, like Meta and Mistral, which also promote open-source AI models.
The rapid advancements in generative AI have spurred intense competition among tech giants, including OpenAI, backed by Microsoft, and Google under Alphabet, leading to massive investments in AI capabilities.
Musk founded xAI roughly 18 months ago, inspired by the rapid success of ChatGPT, with a goal to create the most “truth-seeking” AI. Known for his ventures at Tesla and SpaceX, Musk has been deeply involved in AI development for years. He co-founded OpenAI in 2015 but left in 2018 and has since taken legal action against the ChatGPT maker twice this year.
xAI claims to have built the world’s largest data center in Memphis, Tennessee, where it is advancing Grok, its AI model. Grok, which powers an AI chatbot, is currently accessible exclusively through Musk's social platform X.
The scale of a data center directly influences how quickly a company can train AI models. Musk announced on Monday that xAI plans to expand the Memphis center’s capacity from 100,000 graphics processing units (GPUs) to 200,000. Nvidia CEO Jensen Huang recently praised the rapid development of the Memphis data center, describing it as “easily the fastest supercomputer on the planet” in a podcast interview.
