How Sam Altman stormed Washington to set the AI agenda

NYT News Service

FILE -- Sam Altman (right), OpenAI's CEO, and Sen. Richard Blumenthal, D-Conn., at an hearing on Capitol Hill, Washington, 16 May 2023.

Weeks after

OpenAI

Release its

ChatGPT

Sam, the chatbot from last year.

Altman

The CEO of the artificial-intelligence startup launched a lobbying campaign in Washington.

He demonstrated ChatGPT during a breakfast in the Capitol with over 20 legislators. He called for

Außerdem

To be regulated privately in meetings with Republican or Democratic congressional leaders. Altman has spoken with more than 100 members of Congress as well as Vice President Kamala Harris, Cabinet members and other officials at the White House.

White House

According to legislators and the Biden Administration,

"It is so refreshing," said Sen. Richard Blumenthal (D-Conn.), who was also chair of the panel that conducted an AI hearing featuring Altman last month. "He was willing and able to do it." The spotlight has been avoided by technology chief executives when it comes to government regulators and legislators. In recent years, it took subpoenas to force Mark Zuckerberg from Meta, Jeff Bezos from Amazon, and Sundar Piichai from Google to appear before Congress. Altman, who is 38 years old, has been running toward the spotlight and grabbing the attention of legislators in a manner that has warmed icy attitudes towards Silicon Valley companies. He's organized meetings and has jumped at an opportunity to testify during the Senate hearing last month. Instead of protesting regulation, he invited lawmakers to impose comprehensive rules to hold technology accountable. Altman also took his AI show on tour, delivering the same message in 17 cities across South America, Europe Africa and Asia. He has recently met with President Emmanuel Macron from France, Prime Minister Rishi Sunak from Britain, and Ursula von der Leyen the president of European Commission.

Altman, in a Senate hearing last month, said: "We believe that regulatory intervention by government will be crucial to mitigate the risk of increasingly powerful models." His charm offensive has placed him in a position of significant influence. Altman's early engagement with lawmakers is helping to shape the debate about AI governance and educating Washington as to the complexity of the technology. He is taking a leaf out of the recent past to avoid the pitfalls of social media companies that are constantly targeted by lawmakers and pave the path for AI. His actions could help OpenAI cement its position as a leader in the debate over AI regulation. Microsoft, Google and IBM have drawn battle lines over proposed rules, and they differ in how much government intervention they want to see. These fissures led to other tech leaders pleading their case with the Biden Administration, members of Congress and international regulators. Altman's plan seems to be working so far. Altman has been a trusted educator and advisor to U.S. legislators. He gave a ChatGPT briefing to members of the Senate Select Committee on Intelligence, and the House AI Caucus last month. He has proposed a new independent agency to regulate AI, the licensing of technology and safety standards.

Mark said, "I respect Sam a great deal."

Warner

, D.Va., hosted Altman and more than a dozen senators for dinner last month.

It is not clear how long this goodwill will last. Some lawmakers warned against relying too heavily on Altman or other tech leaders for information on the explosion in new AI technologies. Josh Hawley was the top Republican at the Senate hearing. He said, "He seems different and it was nice to hear him testify." "But I do not think that we should be too complimentary about his company at this time." OpenAI stated that it hoped to bridge the knowledge gaps between Silicon Valley, Washington and AI by learning from past mistakes in the tech industry. Anna Makanju is OpenAI's director of public policy and she leads a team of five experts. She said, "We do not want this to be a repeat of previous technological revolutions." Altman "knows this is an exciting time, so he makes sure to attend as many meetings as possible." Altman has warned about the potential dangers of AI for many years, while also praising the technology. While leading the startup incubator Y Combinator in 2015, he founded OpenAI along with Elon Musk and other Tesla executives. In a blog posting at the time, he wrote that governments should regulate AI's most powerful tools. In an ideal world regulation would slow the bad guys down and speed up good guys, he wrote. Makanju says Altman is a long-time believer that it's better to engage regulators early. Makanju explained that in 2018, OpenAI's mission statement stated its commitment to safety. This implied that regulators would be involved. Ilya Sukseker, the chief scientist of the company, was sent to legislators in 2021 when DALL-E, a tool for AI that generates images based on text commands, was released. Altman spoke at a breakfast organized by the Aspen Institute in Washington, DC. He answered questions, and he previewed GPT-4 - OpenAI's newest AI engine. Altman said that GPT-4 was built with improved security features. Altman's candor regarding AI risks has surprised some legislators. Altman told Rep. Ted Lieu (D-Calif.) in a March meeting at OpenAI offices that AI could have a devastating impact on the labor force, reducing it from five to one day. "He is very direct," said Lieu who has a degree from the Department of Computer Science. Altman returned to Washington in early May, for another White House meeting. This time he met Harris and the CEOs from Microsoft, Google and Anthropic. He also discussed with Chuck Schumer, D.Y., majority leader, his concerns and ideas about China's AI development. Altman made a second appearance with legislators in mid-May. He began with a dinner at the Capitol hosted by Rep. Mike Johnson (R-La.) and Lieu. The event included 60 members of the House. He entertained the audience for two and a half hours with ChatGPT, while serving roasted chicken, potatoes, and salad. According to Lieu, he entered the ChatGPT prompt on a large screen: "Write a law naming a new post office after Representative Ted Lieu." As a second challenge, he asked the audience to "write a speech for Mike Johnson to introduce the bill." Lieu stated that the answers were convincing and elicited chuckles from the audience. Altman spoke at a Senate hearing the next day about AI's dangers. Altman presented a number of regulatory ideas, and he supported legislators' proposals. He also endorsed Blumenthal's idea for consumer risk labels to be placed on AI tools similar to food nutrition labels. Blumenthal stated, "I am so used to hearing witnesses try to convince us by using talking points." "The difference is that Sam Altman has a conversation with us." Altman informed the Senate Intelligence Committee about the security risks of AI after the three-hour hearing. Microsoft has also invested $13 billion into the startup through a partnership with OpenAI. Brad Smith, Microsoft president, stated that he and Altman exchanged feedback on drafts for memos and blogs. Smith stated that the companies coordinated their messaging in advance of the White House event. Smith said that "any day we can support each other in a real way is a great day, because we are trying to accomplish something together." OpenAI was accused by some researchers and competitors of having too much influence on debates about AI regulations. Altman's plans for licensing and testing would benefit established AI companies, according to Marietje Schnaake, former member of the European Parliament and fellow at Stanford's Institute for Human-Centered Artificial Intelligence. Schaake stated that Altman is not just an expert but also a stakeholder.