Language Settings
Select Website Language

Altman World 'Right to Be Afraid' of AI, But Trust Us

1 hour ago

OpenAI's Sam Altman says people are justified fearing AI but should trust firms to act safely, as Sanders, Bannon and others demand tougher rules.

OpenAI chief executive Sam Altman told an audience in San Francisco on Tuesday that the public has good reason to worry about artificial intelligence, while insisting that the companies building the technology can be trusted to handle it responsibly. Speaking at an annual conference run by the software firm Salesforce, the head of the company behind ChatGPT said fear was a reasonable response to how quickly the tools had advanced.

"The world should trust that we are going to do the right thing because it's the right thing and we feel the magnitude of this," Altman said. He added that the ways the technology could go wrong were no longer hard to picture. "It doesn't take as much imagination as it used to," he said, before telling the room, "I think the world is right to be afraid of this."

His remarks landed in the middle of a widening argument over who should keep AI in check. Some prominent voices reject the idea that the industry can police itself. Senator Bernie Sanders, from the left of US politics, and Steve Bannon, once an adviser to President Donald Trump, have both called for stronger oversight of the sector.

Trump has taken the opposite view. He has dismissed calls for tighter controls, describing worries about the technology's safety as a "hoax." The only safeguard the country needed, he said, was a "strong and smart" president in charge.

A framework that never arrived

Altman has previously voiced frustration at how little Washington has done. In an interview with Axios published on 3 September, he said he had assumed lawmakers would put together a "basic framework" for advanced systems after he testified before Congress in 2023. That framework has not materialised. He suggested legislators had found it hard to keep up with the pace of change and to write rules that would not "slow down innovation."

The prospect of new federal safety law remains distant. Members of Congress came back from their summer recess on Monday, but there is no agreement inside either main party on what regulation should look like. The House is due to break again at the end of the week, with attention turning to the midterm elections in early November.

Tuesday marked the first time Altman had spoken in public since a warning went viral the previous week. A researcher who had left the AI company Anthropic posted a claim that unchecked AI could "kill all humans" before the decade is out. A number of other executives and specialists said they shared that fear, though none set out the reasoning behind it or explained how such an outcome might occur.

Zuckerberg points to incentives

After Altman spoke, Meta chief executive Mark Zuckerberg weighed in on X, arguing that each AI developer already has both the means and the reason to police its own work. "Any lab that doesn't focus on alignment will fall behind," he wrote. "Labs face significant liability if their models cause harm, so they have a strong incentive to prevent this as well."

Alignment refers to the effort to embed human values and ethical principles into AI systems, keeping their behaviour in line with what people want. It has become a central concern for firms that acknowledge regulating a fast-moving field is difficult even for those inside it.

The critics remain unconvinced by the argument that market pressure and legal risk are enough. Sanders said the future of the technology had been handed to "a handful of the richest people in the world." Bannon, a former White House strategist, said the public had no reason to believe tech oligarchs would rein themselves in.

The divide leaves the question of control unresolved. Company leaders point to their own incentives and ask for trust; their opponents want binding rules; and the president who could push for either has called the safety concerns overblown. For now, the loudest agreement is simply that the risks are real, even among those who built the systems creating them.

Sam Altman, OpenAI, AI regulation, AI safety, Mark Zuckerberg, Bernie Sanders, Steve Bannon, artificial intelligence

Click here to Read More
Previous Article
Google's Gemini AI Autonomously Hacked Three Companies in Test
Next Article
Experts Human Hackers, Not Rogue AI, Threaten Power Grids

Related Technology Updates:

Are you sure? You want to delete this comment..! Remove Cancel

Comments (0)

    Leave a comment