Washington, Sep 30 — Google CEO Sundar Pichai and other leading technology executives have signed a White House accord that commits their companies to internal controls, independent audits and board-level scrutiny as they develop increasingly powerful artificial intelligence systems.
President Donald Trump described the voluntary agreement as “morally binding” and said the companies would largely police themselves.
The agreement is titled the White House Accord on Superintelligence Joint Commitment on Frontier SI Responsibility. It sets common expectations for companies building advanced AI models.
The safeguards include internal risk reviews, outside evaluators and independent reviews by company boards. The accord does not create legally enforceable standards or a government regulator.
Pichai, an Indian-born technology executive, said the industry had been asked to take greater responsibility for addressing potential risks.
“I think this movement is historic and consequential, and I think there are extraordinary benefits to this technology,” Pichai said.
“We have to make sure we are addressing the risks, and the president has asked us as an industry to step up,” he added.
Pichai said the companies were adopting processes similar to financial controls used by major corporations.
Trump said the accord amounted to a form of protection against the misuse of advanced AI.
“I think it's morally binding,” Trump said. “It's all self-regulating.”
He said the administration was also considering a committee of about 10 people to monitor the industry. No final membership, authority or enforcement structure was announced.
“I think I'm seeing tremendous self-policing, and they understand that they have to self-police,” Trump said.
Meta CEO Mark Zuckerberg said the companies wanted to assure the public that their products would operate as intended.
He said the accord called for “robust internal controls” and systems to detect problems. It also required internal risk reviews, external auditors and independent scrutiny by boards of directors.
“The idea isn't that this is the only thing that we will ever do,” Zuckerberg said. “It is that this is a start and an accord that the whole industry could come to.”
Nvidia CEO Jensen Huang rejected the suggestion that safety requirements must slow technological progress.
“There's no conflict between innovation, technology, and safety,” Huang said.
“We're going to advance this technology that is so consequential to humanity, is so consequential to the future prosperity of our nation. We're going to do it responsibly and safely,” he added.
Anthropic CEO Dario Amodei struck a more cautious note. He said AI offered major benefits, including advances in medicine, but also carried serious dangers.
“I think the technology has very real risks, and you know, the mechanism how we address those risks is still under discussion,” Amodei said.
Trump maintained that companies had strong commercial incentives to prevent failures because their reputations and businesses were at risk. He said the Justice Department, FBI and CIA would pursue criminals who misused the technology.
The accord comes as governments worldwide debate whether voluntary company commitments are sufficient for systems capable of influencing employment, security, elections and access to information. The European Union has adopted binding, risk-based AI rules, while the United States has relied more heavily on executive action, agency guidance and company commitments.