The former OpenAI employees are asking the main law agents in California and Delaware who prevent the company from changing the control of its artificial intelligence technology Of a non -profit beneficial organization to a profit for profit.
They are concerned about what happens if the Chatgpt manufacturer fulfills their ambition to build an AI that surpasses humans, but is no longer responsible to their public mission to safeguard that technology to cause serious damage.
“Ultimately, I am worried who possesses and controls this technology once it is created,” said Page Hedley, former policy and ethical advisor at OpenAI, in an interview with Associated Press.
Backed by three winners of the Nobel Prize and other defenders and experts, Hedley and nine other open workers sent a letter this week to the two general state prosecutors.
The coalition asks California Rob Bronta attorney to the Attorney General of Delaware, Kathy Jennings, both Democrats, who use their authority to protect Openai’s charitable purpose and block their planned restructuring. OpenAi joins Delaware and operates from San Francisco.
Openai said in response that “any change in our existing structure would be in service to ensure that the general public can benefit from AI.” He said that his for profit will be a public benefits corporation, similar to other AI laboratories such as the Anthrope and the technological billionaire Elon Musk, except that Openai will still preserve a non -profit arm.
“This structure will continue to ensure that as the for profit is successful and grows, so does the non -profit organization, which allows us to reach the mission,” the company said in a statement.
The letter is the second request to state officials this month. The last wine of a group of labor leaders and non -profit organizations focused on protecting the billions of OpenAi dollars of charitable assets.
Jennings said the past fall that “would review any transaction of this type to ensure that the interests of the public are properly protected.” The Bronta office looked for more Openai information at the end of last year, but has said that it cannot comment, even to confirm or deny if you are investigating.
Openai co -founders, including the current CEO Sam Altman and Musk, originally began as a non -profit research laboratory on a mission to safely build what is known safely as Artificial general intelligence, or agifor the benefit of humanity. Almost a decade later, Openai has reported its market value as $ 300 billion and counts 400 million weekly chatgpt users, its flagship product.
Operai already has a subsidiary for profit, but faces a series of challenges to convert its central government structure. One is a Almizcle demandwho accuses the company and Altman of betraying the founding principles that led the CEO of Tesla to invest in the beneficial organization.
While some of this week’s letter signatories support Musk’s demand, Hedley said that others are “understandably cynical” because Musk also directs his own rival company.
The signatories include two winning economists of the Nobel, Oliver Hart and Joseph Stiglitz, as well as the Pioneers of AI and Computer scientists Geoffrey Hintonwho won last year Nobel Prize in Physicsand Stuart Russell.
“I like Openai’s mission to” ensure that artificial general intelligence benefits all humanity “, and I would like to execute that mission instead of enriching its investors,” said Hinton in a statement on Wednesday. “I am happy that there is an effort to keep OpenAi to his mission that does not involve Elon Musk.”
Conflicts about Openai’s purpose have done over the San Francisco Institute, contributing to Musk to resign in 2018, Altman’s Short duration expulsion in 2023 and other high profile outputs.
Hedley, a training lawyer, worked for OpenAI in 2017 and 2018, a time when the non -profit organization was still sailing for the best ways to administer the technology I wanted to build. As recently as 2023, Altman said Advancez AI was promised, but also warned of extraordinary risks, from drastic accidents to social interruptions.
In recent years, however, Hedley said he observed with concern as Openai, driven by Chatgpt’s success, was increasingly reducing corners in security tests and hurrying new products to get ahead of commercial competitors.
“The costs of these decisions will continue to increase as technology becomes more powerful,” he said. “I think that in the new structure that Operai wants, the incentives to hurry to make those decisions will rise and there will be no one who can tell them that they do not, tell them that this is not right.”
The software engineer Anish Tondwalkar, a former member of the OpenAI technical team until last year, said that an important guarantee in the non -profit letter of Openai is a “stop and assistance clause” that orders OpenAi to withdraw and help if another organization is approaching the achievement of the best AI of the AI.
“If Openai is allowed to become a for profit, these safeguards and Openai’s duty with the public can disappear overnight,” Tondwalkar said in a statement on Wednesday.
Another former worker who signed the letter puts him more bluntly.
“Operai can build technology that could kill us all,” said Nisan Stiennon, an AI engineer who worked in Openai from 2018 to 2020. “It is Openai credit that is controlled by a non -profit organization with a duty for humanity. This duty prevents giving up this control.”
___
Associated Press and OpenAi have A license and technology agreement That allows Openai access to part of the AP text files.
