Elon Musk Tried to Hire Sam Altman for Tesla Before Leaving OpenAI

Court documents from an ongoing lawsuit between Elon Musk and Sam Altman show that Musk tried to recruit Altman to work on artificial intelligence research at Tesla. The job offer included a seat on Tesla's board of directors. This happened just months before Musk left OpenAI's board in February 2018, according to emails and testimony presented in federal court.

Shivon Zilis, a Tesla executive, acted as a go-between during these conversations. Court documents show she and other Tesla leaders discussed plans to start a rival AI lab at Tesla, possibly led by Altman or Demis Hassabis, a co-founder of DeepMind.

What the Two Sides Wanted

Musk's lawsuit claims that Altman and OpenAI president Greg Brockman wrongly took control of OpenAI. Court evidence shows Musk had his own plan: he wanted to absorb OpenAI into Tesla and run it himself.

Both Musk and OpenAI agreed in 2017 that the organization needed to become a for-profit company instead of staying as a nonprofit. Musk created a legal structure called a Public Benefit Corporation and pushed for a majority stake and full control. When OpenAI refused, negotiations fell apart.

In January 2018, Musk told OpenAI leaders the company was heading toward failure without merging with Tesla. A month later, he stepped down from OpenAI's board. He said he wanted to focus on building advanced AI at Tesla instead.

The Money at Stake

The financial numbers in this dispute are large. Musk says he invested 38 million dollars in OpenAI before it became a private company. Court documents now say OpenAI is worth more than 800 billion dollars, with some reports putting the value at 852 billion dollars as the company prepares to go public.

To put that in perspective, those sums far exceed compensation disputes Musk has faced elsewhere. A Delaware court previously rejected Musk's 55.8 billion dollar Tesla pay package, calling it "250 times larger" than comparable executive compensation. Tesla is now moving its headquarters to Texas to change which state's courts oversee future disputes.

Why This Matters

This case shows the intense competition for AI talent and control that played out during the early years of commercial artificial intelligence, around 2017 and 2018. Major technology companies were building their own AI labs and competing hard to hire the same small group of experienced leaders.

History suggests this pattern has played out before. When cloud computing—the practice of renting computing power and storage over the internet rather than owning servers—emerged in the mid-2000s, Amazon, Microsoft, and Google fought similar battles over talent, structure, and control as they built their cloud divisions.

Both Musk and OpenAI's leaders had different ideas about how to develop advanced AI. Musk wanted to build AI directly into Tesla's cars and energy business. OpenAI's leadership wanted to keep research somewhat independent while partnering with outside investors for money.

What Happens Next

The trial is taking place in federal court in Oakland, California. Jury selection began in late April. The judge hopes to have a jury decision by late May about whether OpenAI broke any promises made to Musk. Musk reportedly tried to settle before trial started, though the settlement terms were not made public.

Court papers also show the judge has ordered both sides to stop using disappearing message apps like Signal for business communications. This is to make sure important documents are preserved as evidence.

A Broader Picture

The details revealed in court show how Musk and OpenAI's leaders thought about organizing large-scale AI research back then. Musk believed that putting an AI lab inside Tesla would have advantages—research could move faster from idea to product because Tesla already had engineers, factories, and expertise in manufacturing.

That approach had downsides too. Focusing AI research within one company's needs might have made it harder to hire and keep top researchers who prefer working on pure science. OpenAI went a different direction: it partnered with Microsoft and focused on building language models like GPT, instead of tying research tightly to one company's products.

The core question in this case is bigger than just what Musk and Altman want. It touches on how courts should handle disagreements involving organizations that start as nonprofits but later become profitable businesses. As more technology companies mix research and business goals, these kinds of disputes may become more common.