Blumenthal: AI deepfake ‘one of the more scary moments’ in Senate hearing history
“It was eerie, even creepy," the Connecticut senator said.
TroibNews 
There was an “eerie, even creepy” atmosphere during opening remarks of a Senate hearing on AI Tuesday morning. But the feeling didn’t hit immediately, because no one could tell artificial intelligence was impersonating a U.S. senator.
“If you closed your eyes at the beginning of the hearing, you couldn't have told that we were playing a voice clone of myself in introducing the hearing,” Sen. Richard Blumenthal (D-Conn.) said on MSNBC Wednesday morning. “It was eerie, even creepy.”
Blumenthal commenced the hearing, which featured testimony from OpenAI CEO Sam Altman, with a faked voice recording that was written by ChatGPT and vocalized by an audio application trained on his Senate floor speeches. The recording outlined the reason for the hearing and warned of when “technology outpaces regulation,” possibly leading to misinformation and the exploitation of personal data.
The real Blumenthal delivered a severe warning after the demonstration: What if, he said, the recording had been used to fake an endorsement of Ukraine surrendering, or of Russia President Vladimir Putin?
“No one, in fact, could tell that it was a clone. I think that it is one of the more scary moments in the United States Senate hearing history,” Blumenthal said on MSNBC.
