‘President of the Unites States’: White House’s AI accord includes spelling error

The White House misspelled “United States” under the signature of Donald Trump on an accord he signed with AI executives on Tuesday.

At the bottom of the accord on “Super Intelligence”, the term Trump has used to rebrand artificial intelligence, the president’s signature appeared as “President of the Unites States”.

It is unclear how the misspelling occurred or whether it will appear in the Federal Register, the database that contains official presidential actions and other government documents.

Trump and six tech executives signed the document voluntarily committing to safely developing AI to address widespread concerns, sparked by industry leaders who warned the technology is evolving too rapidly to control.

Trump has largely brushed aside concerns about AI harming humans and called for companies to continue developing the technology so the US can become the global leader in AI. Even while signing the accord on safety, the US president praised the business executives for growth.

Trump posted the document containing the misspelling on Truth Social , and the White House Communications Team emailed it out shortly after it was signed.

The Bloomberg News correspondent David Gura appeared to be the first to notice the spelling error.

The accord on artificial intelligence outlined four layers of “controls and audits” that companies will implement to ensure AI capabilities operate as intended. That includes “robust” internal controls and partnering with an “independent external auditor or evaluator”.

Potential internal controls could include empowering a team to monitor and detect that AI is operating correctly and creating monitors to ensure AI technologies will not hack or access technical systems in cybersecurity, biosecurity and chemical threats.

External monitors would also check to make sure the AI is operating as intended and companies could designate a committee of the board of directors to oversee reports from teams operating the controls of the AI.

The accord is the first step in addressing concerns, raised by industry experts in mid-September.

Americans have become increasingly skeptical of AI. A recent New York Times/Siena College poll found that 61% of voters opposed the construction of datacenters, while an SSRS/CNN poll found that 71% of Americans believed the government wasn’t doing enough to regulate the growth and development of AI.

While Tuesday’s document is only an accord, it left open the possibility of codifying the framework into formal regulations or laws – something Trump has repeatedly rejected.

“I think I’m seeing tremendous self-policing,” Trump told reporters outside the White House on Tuesday. “And they understand that they have to self-police.”

The six executives who signed the accord are Dario Amodei from Anthropic; Elon Musk from SpaceXAI; Greg Brockman from OpenAI; Jensen Huang from Nvidia; Mark Zuckerberg from Meta; and Sundar Pichai from Google.

本文の著作権はThe Guardianにあります。