one year on
OpenAI CEO Sam Altman tells Senate AI regulation is essential, warns technology needs safeguards
At a Senate Judiciary subcommittee hearing, Altman endorses licensing for AI models above a threshold of capabilities as lawmakers across two panels show unusual bipartisan willingness to regulate.
OpenAI CEO Sam Altman today urged a Senate Judiciary subcommittee to impose licensing requirements on the most powerful AI systems, calling regulatory intervention “critical” to manage what he described as a technology that could become a transformative “printing press moment” if handled right — but that could go wrong without safeguards.
Speaking alongside IBM’s vice president and chief privacy and trust officer Christina Montgomery and NYU Professor Emeritus Gary Marcus, Altman said he fears AI’s potential to manipulate voters and spread disinformation, especially ahead of the 2024 election. He endorsed a regime that would license companies developing models above a certain capability threshold, requiring testing before release.
Senator Richard Blumenthal opened the hearing by playing a deepfake audio of his own voice — generated by ChatGPT — reading AI-written remarks that argued for regulation. Blumenthal noted the same technology could have produced an endorsement of Ukraine’s surrendering or Vladimir Putin’s leadership.
The tone of the hearing was notably cooperative, with lawmakers across the aisle appearing eager to act. Altman’s appearance followed a bipartisan dinner with more than 60 House members the night before, where he demonstrated ChatGPT. Separately, the Senate Homeland Security Committee held a simultaneous hearing on AI in government. The week’s prevailing sentiment: the AI chief himself is asking for rules, and Congress is listening.
The record
As a witness, Marcus, a critic of AI hype, testified alongside Altman and Montgomery.
IBM’s vice president and chief privacy and trust officer warned against a new era of 'move fast and break things' but said there is no need to slam the brakes on innovation.
After a bipartisan dinner with Altman, Khanna said the CEO's most helpful contribution was 'ramping down the hype' and stressing AI is a tool, not a creature.
One year later — open only if you can handle spoilers
The cooperative optimism of May 2023 soon cooled. By 2024, the Senate had not passed comprehensive AI legislation, and Altman faced growing criticism from both safety advocates and competitors. The licensing idea Altman endorsed never became law.