A voice rings out in the hearing room.

It's Senator Blumenthal's voice - but he's not speaking.

The senator has played an opening statement generated by AI and written by ChatGPT, after instructions from him to generate opening remarks on the future of regulation before a judiciary subcommittee based on his past speeches in Congress.

He says he does this to show the potential harm of artificial intelligence.

"My playing this might strike you as humourous," he says.

But he warns: "What if he had provided an endorsement of Ukraine surrendering?"

The hearing is already getting interesting.

Image source, None

The chair, Connecticut Senator Richard Blumenthal, is now speaking.

He introduces the hearing by saying its "goal is to demystify and hold accountable those new technologies to avoid some of the mistakes of the past".

Stream live at the top of the page.

Zoe Kleinman
Technology editor

Image source, Reuters

I’ve watched a few tech leaders get a grilling from US congress - some of them handle it better than others but none of them get an easy ride.

The questions can be ferocious and they can also be difficult to follow. In some sessions representatives get a maximum of six minutes to ask each individual question and some of them seem determined to make the most of every single second.

Then there are the varying levels of understanding of the tech sector and how it works.

One Senator memorably didn’t seem to know how Facebook had made its vast amounts of money when it was Meta boss Mark Zuckerberg’s turn to testify.

“Senator, we run ads,” Zuckerberg famously replied, with a hint of a smile.

Stick around long enough and you might find yourself marvelling at the thought processes of some of the lawmakers, but make no mistake, this is still a lion’s den.

The Financial Times described TikTok CEO Shou Zi Chew’s recent five hour-long appearance as a “bipartisan pummelling”.

Very few people understand the inner workings of ChatGPT, but today is the US’s opportunity to press its creator on its impact, its potential, and what is being done to keep it safe. And the world will be listening.

James Clayton
North America technology reporter

I’m expecting a number of lines of attack today including:

• Political bias - Many Republicans believe ChatGPT is too “woke”. Expect lots of examples of this from conservative senators.
• Jobs - I expect politicians from both sides will raise this. Altman himself admits this technology will lead to the loss of millions of jobs - but will also create “better” ones. He’ll be pushed on this.
• Cheating - GPT-4 can pass law and medical exams. Can we trust the sanctity of exams with this technology around?
• Accuracy - For anyone who’s used GPT-4, you’ll know that its answers can be incredibly useful and also totally wrong. GPT-4 told me that this hearing was in fact due to take place yesterday, for example. There are worries the technology could lead to more disinformation.
• Regulation - Pretty much everyone agrees AGI (artificial general intelligence) needs to be regulated. The question is how and by whom. OpenAI want to get ahead of this, they’d rather help make the rules.

Marianna Brady
Live reporter, Washington DC

Hello and welcome to our live coverage.

The founder of ChatGPT will testify before the Senate subcommittee on privacy today in Washington DC.

Sam Altman, CEO of OpenAI, which created the famous chatbot, will face lawmakers on a series of issues relating to artificial intelligence, or AI.

The field of AI has been under sharp scrutiny from experts in the tech community and lawmakers around the world - and there are still many questions to be answered.

We expect senators to grill Altman on how big of a threat AI is, and how to best regulate it.

Today’s hearing - due to begin at 10:00 local time (15:00 BST) - will be Altman’s first appearance before lawmakers.

We'll be streaming the hearing live at the top of the page.