The owners and managers of the main artificial intelligence companies in the United States signed a joint security commitment with President Donald Trump on Tuesday, as part of a meeting held at the White House with industry leaders to respond to growing warnings about the risks of this technology. The announcement was made on Tuesday during a lunch convened by Trump and the Speaker of the House of Representatives, Mike Johnson (ABC7).

Sundar Pichai (Google), Dario Amodei (Anthropic), Mark Zuckerberg (Meta), Greg Brockman (OpenAI), Elon Musk and Jensen Huang (Nvidia) signed the document with Trump, as confirmed by Ynet News, which detailed the content of the agreement published by the president himself on his account on the social network Truth Social. In addition to the executives, representatives and executives from Palantir and Amazon, among others, attended (ABC7).

The document, described by Trump as an industry self-regulatory agreement, establishes four layers of control and oversight for signatory companies. Each company committed to implementing real security controls in its most powerful AI models, bringing in external auditors or evaluators to make independent assessments on whether its control and monitoring systems are operating properly, and being accountable to an independent board committee that oversees the reports of internal and external auditors (The Hindu, The Straits Times).

The central objective, according to the agreement, is to prevent these systems from being used for cyberattacks or biological and chemical threats. The text calls for internal controls to monitor the capabilities and alignment of models during their training and deployment in areas such as cybersecurity, biosecurity and chemical threats, ensuring that models "do not hack or access technical systems unintentionally". It also calls for an internal team to verify that all controls, monitoring and detection are working as expected (KELO/AP, Reuters).

Trump called the compromise "morally binding" and proposed the creation of a committee of about 10 people to "oversee the whole picture," with a person responsible for the deal to be appointed in the coming days after consulting with industry. Mark Zuckerberg said the pact includes "a set of principles and commitments to build robust internal controls" and that "all of our boards will independently review auditors' reports" (HuffPost, ABC7).

The announcement comes amid an intense debate in Washington about the future of artificial intelligence: Johnson had proposed that Trump and the executives "enter a big room, close the door and resolve the matter," while the White House resists pressure to impose regulations that could slow the technological race against China (POLITICO, FMT).

Sources: ABC7/Associated Press · Ynet News · The Hindu/Reuters · The Straits Times/Reuters · HuffPost · CNBC · Washington Examiner · KELO/AP. Image credit: Molotov News.