On August 6, the District Court for the Northern District of California in United States accepted a new indictment case, in which Musk sued OpenAI and its co-founders Sam Altman and Gregory Brockman for violations of federal law, fraud, breach of fiduciary duty, unfair competition, and other grounds.
Previously, Musk had sued them on March 2 of this year, and later gave up when the trial was about to start on June 12.
It is estimated that Lao Ma is going to play really this time, and the lawsuit alone has written 83 pages, and some of the core reasons are directly to the point of OpenAI.
Background to the prosecution
In 2014, Google's acquisition of DeepMind triggered Musk's concerns about AI, fearing that Google would monopolize the technology.
In 2015, Musk approached Altman to discuss building a non-profit, open AI organization dedicated to developing secure and open AI technologies for the benefit of all mankind.
OpenAI was officially established in late 2015, and Altman promised that OpenAI would operate on a non-profit basis, open source the technology, avoid technology concentration, and not operate for personal or corporate gain.
Since 2015, Musk has provided financial support, advice, and used his reputation and connections to recruit top scientists and engineers for OpenAI.
This period included investments of more than $44 million, including $15 million in 2016, $20 million in 2017, and other fragmented investments, while helping OpenAI recruit key talent such as Ilya Sutskever, the world's top AI scientist who has since left the company.
Beginning in 2017, Altman and Brockman began to try to transform OpenAI into a for-profit enterprise, but were harshly rejected by Musk. Musk asked them to continue to adhere to OpenAI's non-profit nature and mission, otherwise he will stop supporting OpenAI's various resources.
Under pressure from Musk, Altman and Brockman have temporarily abandoned the idea, but they have not given up on their plans to commercialize OpenAI.
In 2018, Musk left OpenAI because of a serious disagreement with Altman about OpenAI's development philosophy and wanted to take control of OpenAI but was rejected by the board of directors.
The core cause of action
Promise fraud: Musk accused Altman and Brockman of inducing him to fund and support OpenAI through false promises starting in 2015. They claim that OpenAI will be a non-profit organization dedicated to developing secure and open-source AI for the benefit of humanity, but in reality they planned from the beginning to transform OpenAI into a for-profit enterprise and profit from it. Musk poured a lot of money into OpenAI based on these promises, but the defendants did not fulfill their promises and instead violated OpenAI's non-profit mission by closing the technology and pursuing commercial interests.
Constructive Fraud: OpenAI, Altman, and Brockman had a fiduciary relationship with Musk as a charitable organization and a person seeking donations on behalf of a charitable organization, but they violated their fiduciary duty by misleading Musk and concealing their true business intentions, constituting constructive fraud.
Breach of express contract: Musk claims that he and Altman reached an agreement in 2015 through a series of written communications to co-create OpenAI as a nonprofit entity to develop leading-edge AI/AGI technology.
Under the agreement, Musk will provide funding, advice and recruit talent, while Altman promises that OpenAI will be non-profit and open source the technology. But in 2023, Altman and OpenAI violated the agreement, including failing to disclose research results, licensing the technology exclusively to Microsoft, and trading itself.
Violation of security, public interest: From 2023 onwards, OpenAI's behavior has become more worrying, and it has begun to conceal its research and development results, for example, for models such as GPT-4 and GPT-4o, OpenAI has not disclosed details such as its architecture, hardware, training methods, and training calculations.
At the same time, OpenAI exclusively licensed these technologies to Microsoft, concentrating them in the large for-profit company, violating its original promise that the technology would be primarily open source and serve the public interest.
False advertising: Musk accused Altman and Brockman of engaging and supporting OpenAI through false marketing and propaganda that led him into believing that OpenAI is a nonprofit dedicated to developing secure and open-source AI technologies. But in reality, OpenAI's actions did not match their propaganda and constituted false advertising, violating the Lanham Act and the California Business and Profession Principles.
Aiding and abetting breach of fiduciary duty: Musk believes that OpenAI's for-profit entity aided and supported Altman, Brockman, and OpenAI knowing that they were breaching his fiduciary duty.
For example, these entities helped Altman and Brockman move OpenAI's intellectual property and employees into for-profit structures to achieve their commercial interests, rather than furthering OpenAI's nonprofit mission, and this behavior constitutes aiding and abetting breaches of fiduciary duty.
Musk said that he asked the court to order the defendant to pay corresponding compensation, including compensatory, consequential and statutory damages and demand the payment of corresponding interest, and at the same time determine the validity of OpenAI's license agreement with Microsoft, as well as whether GPT-4, GPT-4o and other models constitute AGI and whether it is beyond the scope of Microsoft's license agreement.