Elon Musk sues OpenAI and Sam Altman for breach of contract

Elon Musk has sued OpenAI and its chief executive Sam Altman for breach of contract, alleging they have compromised the start-up’s original mission of building artificial intelligence systems for the benefit of humanity.

In the lawsuit, filed to a San Francisco court on Thursday, Musk’s lawyers wrote that OpenAI’s multibillion-dollar alliance with Microsoft had broken an agreement to make a big breakthrough in AI “freely available to the public”.

Instead, the lawsuit said, OpenAI was working on “proprietary technology to maximise profits for literally the largest company in the world”.

The legal fight escalates a long-running dispute between Musk, who has founded his own AI company, known as xAI, and OpenAI, which has received a $13bn investment from Microsoft.

Musk, who helped co-found OpenAI in 2015, said in his legal filing he had donated $44mn to the group, and had been “induced” to make contributions by promises, “including in writing”, that it would remain a non-profit organisation.

He left OpenAI’s board in 2018 following disagreements with Altman on the direction of research. A year later, the group established the for-profit arm that Microsoft has invested into.

Microsoft’s president Brad Smith told the Financial Times this week that while the companies were “very important partners”, “Microsoft does not control OpenAI”.

Musk’s lawsuit alleges that OpenAI’s latest AI model, GPT4, released in March last year, breached the threshold for artificial general intelligence (AGI), at which computers function at or above the level of human intelligence.

The Microsoft deal only gives the tech giant a licence to OpenAI’s pre-AGI technology, the lawsuit said, and determining when this threshold is reached is crucial to Musk’s case.

The lawsuit seeks a court judgment over whether GPT4 should already be considered to be AGI, arguing that OpenAI’s board was “ill-equipped” to make such a determination.

The filing adds that OpenAI is also building another model, Q*, that will be even more powerful and capable than GPT4. It argues that OpenAI is committed under the terms of its founding agreement to make such technology available publicly.

“Mr Musk has long recognised that AGI poses a grave threat to humanity — perhaps the greatest existential threat we face today,” the lawsuit says.

“To this day, OpenAI, Inc.’s website continues to profess that its charter is to ensure that AGI ‘benefits all of humanity’,” it adds. “In reality, however, OpenAI, Inc. has been transformed into a closed-source de facto subsidiary of the largest technology company in the world: Microsoft.”

OpenAI maintains it has not yet achieved AGI, despite its models’ success in language and reasoning tasks. Large language models like GPT4 still generate errors, fabrications and so-called hallucinations.

The lawsuit also seeks to “compel” OpenAI to adhere to its founding agreement to build technology that does not simply benefit individuals such as Altman and corporations such as Microsoft.

Musk’s own xAI company is a direct competitor to OpenAI and launched its first product, a chatbot named Grok, in December.

OpenAI and Microsoft declined to comment. Representatives for Musk have been approached for comment.

In a letter to the company seen by the FT, Jason Kwon, OpenAI’s chief strategy officer, said they “categorically disagree” with the premise of the lawsuit, adding that GPT4 was not AGI and that OpenAI was an independent entity. 

“We believe the claims in this suit may stem from Elon’s regrets about not being involved with the company today,” Kwon wrote.

“Early on, we all agreed on needing a new structure to raise far more capital for talent and compute. We discussed a for-profit entity seriously, but Elon said he needed full initial control and majority equity, and he later suggested merging OpenAI into Tesla. We did not think either approach was right for the mission.” 

The Microsoft-OpenAI alliance is being reviewed by competition watchdogs in the US, EU and UK.

The US Securities and Exchange Commission issued subpoenas to OpenAI executives in November as part of an investigation into whether Altman had misled its investors, according to people familiar with the move.

That investigation came shortly after OpenAI’s board fired Altman as chief executive only to reinstate him days later. A new board has since been instituted including former Salesforce co-chief executive Bret Taylor as chair.

There is an internal review of the former board’s allegations against Altman by independent law firm WilmerHale.