OpenAI CEO Sam Altman—the man behind ChatGPT—issued a stark warning to the Senate Judiciary Committee on Tuesday: Artificial Intelligence can be dangerous.
“If this technology goes wrong, it can go quite wrong, and we want to be vocal about that,” Altman said, separately noting that his “worst fear is we cause significant harm to the world.”
It was an unusual moment for a major tech entrepreneur to so boldly agree with lawmakers, who spent much of Tuesday’s hearing pointing out the potential perils of his product. Most congressional panels on major tech platforms like TikTok and Facebook devolve into hours-long attempts at “gotcha” questions and defensive remarks.
But Altman’s broader message Tuesday was welcoming and actively encouraging regulation on Artificial Intelligence, including his own ChatGPT. Altman said he fears ChatGPT’s potential to manipulate users’ views. He suggested forming a new government agency to govern AI models. And he agreed that AI would take away some jobs—though, he suggested new jobs would be created, too.
“ChatGPT4 will, I think, entirely automate away some jobs. And it will create new ones that we believe will be much better,” Altman said.
“There will be an impact on jobs,” he added. “We try to be very clear about that.”
Altman did push back against some forms of regulation, like granting users transparency into program training data.
Tuesday’s hearing had the potential to be an embarrassment for senators. Past hearings on Big Tech have delivered now-notorious gaffes from the elderly, largely tech-inept members of the Senate, like Sen. Richard Blumenthal (D-CT) asking Instagram to end “finstas.” Sen. Orrin Hatch (R-UT) once asked Facebook CEO Mark Zuckerberg how the platform makes money, to which Zuckerberg awkwardly responded, “Senator, we run ads.”
But lawmakers Tuesday, for the most part, seemed well-prepared on the subject at hand. Some expressed gratitude to Altman not just for his attendance, but for previous conversations the CEO had with lawmakers regarding concerns and regulation. According to Axios, Altman dined with about 60 lawmakers the night before the hearing.
The Senate has shown interest in regulating AI this term. Senate Majority Leader Chuck Schumer (D-NY), who’s piercing together his own proposal, is one of them. Many senators at Tuesday’s hearing expressed their dissatisfaction that Congress had been slow regulating other tech, like social media, leading to a long game of catchup.
With booming AI products like ChatGPT, lawmakers are hoping things will be different. Yet, as the hearing ended Tuesday, it wasn’t immediately clear what action Congress would or could take.
When asked, ChatGPT, which claims to not have any information on current events that occurred after Sept. 2021, would not remark on Altman’s performance before the panel Tuesday.
The Daily Beast also asked ChatGPT if it should be regulated by Congress, as Altman suggested. The AI said that is a “complex and debated topic” and listed a number of pros and cons to regulation while taking no official stance.
“As an AI language model, I don't have personal opinions or beliefs,” ChatGPT remarked. “I can provide information and perspectives on different topics, but I do not hold any specific stance on whether Congress should regulate AI or agree with any particular individual, including Sam Altman, on this matter.”