The US Senate welcomed American technology leaders to Capitol Hill on Wednesday for a closed-door forum on the need to set artificial intelligence safeguards.
Congress is grappling with how to mitigate the dangers of the emerging technology, which has experienced a boom in investment and consumer popularity after the release of OpenAI's ChatGPT.
“This is an important, urgent and in some ways unprecedented moment,” OpenAI chief executive Sam Altman said as he walked into the meeting in Washington.
Politicians want safeguards against potentially dangerous deepfakes which could lead to issues from bogus videos to election interference and attacks on critical infrastructure.
“Today, we begin an enormous and complex and vital undertaking: building a foundation for bipartisan AI policy that Congress can pass,” US Senate Majority Leader Chuck Schumer, a Democrat, said in opening remarks.
“Congress must play a role, because without Congress, we will neither maximise AI’s benefits nor minimise its risks.”
Other attendees included Nvidia chief executive Jensen Huang, Microsoft head Satya Nadella, IBM's Arvind Krishna, Microsoft founder Bill Gates.
Mr Schumer, who talked AI with Mr Musk in April, wanted attendees to talk “about why Congress must act, what questions to ask, and how to build a consensus for safe innovation”.
In March, Mr Musk and a group of AI experts and executives called for a six-month pause in developing systems more powerful than OpenAI's GPT-4, citing potential risks to society.
This week, Congress is holding three separate hearings on AI.
Microsoft president Brad Smith told a Senate Judiciary subcommittee on Tuesday that Congress should “require safety brakes for AI that controls or manages critical infrastructure”.
Mr Smith compared AI safeguards to requiring circuit breakers in buildings, school buses having emergency brakes and planes having collision avoidance systems.
Republican Senator Josh Hawley questioned the need for Wednesday's closed-door session.
“I don't know why we would invite all the biggest monopolists in the world to come and give Congress tips on how to help them make more money,” Mr Hawley said, pointing out that the government has yet to pass any meaningful AI legislation.
Regulators globally have been scrambling to draw up rules governing the use of generative AI, which can create text and generate images whose artificial origins are virtually undetectable.
Adobe, IBM, Nvidia and five other companies on Tuesday said they had signed President Joe Biden's voluntary AI commitments, which require steps such as watermarking AI-generated content.
The commitments, which were announced in July, were aimed at ensuring AI's power was not used for destructive purposes. Google, OpenAI and Microsoft signed on in July.
The White House has also been working on an AI executive order.