It sure seems like the people who have the most to gain from an unregulated tech industry are the ones saying that the concerns over AI are overblown.
Ex-AI researchers sound alarm on AI safety
Anthropic’s head of AI alignment, Evan Hubinger, said in a post that he believes there’s more than a 10% chance AI could “kill all humans” in the next 10 years.
A former artificial intelligence researcher is sounding the alarm on how the technology he helped develop could lead to the demise of humanity. Now, even Anthropic CEO Dario Amodei is saying regulation needs to happen. If only politicians would take these concerns seriously.
Jacob Coxon, an Anthropic employee who resigned from the company on Sept. 8, posted on X about his decision to leave. A former employee of OpenAI, he said that neither company had things under control amid the rapid advancement of AI.
“The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon said on X. “This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible – but I hear the same people express fear privately.”
Adding to the anxiety was a reply from Evan Hubinger, the Alignment Science lead at Anthropic, who gave a grim estimate of the timeline before AI takes over and kills everyone on earth.
“We really do earnestly believe AI could kill all humans,” Hubinger wrote. “I personally think it is >10% within the next decade.”
As if these posts weren’t enough to raise alarms, there are recent revelations about Anthropic intervening in bioweapons creation and the threat of foreign actors using Claude, the company’s chatbot, to create weapons.
According to a recent poll from Pew Research Center, 52% of Americans are “more concerned than excited” about AI’s increased presence in our daily lives. There is a huge, bipartisan grassroots movement against data centers being built in communities. Recent news about Anthropic further proves to the American people that AI is even more of a headache than we initially thought.
Of course billionaires like Trump, Musk aren’t worried about AI
It’s troubling that the tech workers who helped build the AI landscape we see today are saying that things are moving too fast, and it’s equally concerning that the powers that be do not care enough to pass comprehensive regulation on the industry.
“I have concerns that if we don’t win AI, we’re going to be put in a very bad position,” President Donald Trump said on Sept. 10, noting that the greater concern for him is the United States not winning the AI battle with China.
On Sept. 14, he furthered his stance, saying on Truth Social that “the only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!”
Meanwhile, tech trillionaire Elon Musk called the posts a “psyop.” It’s a convenient narrative for someone who makes money off of his own artificial intelligence company.
It sure seems like the people who have the most to gain from an unregulated tech industry are the ones saying that the concerns raised by Coxon and others are overblown.
While I do worry that the concerns raised by these employees teeter on “doomerism,” I think it’s better that we listen to them than avoid taking action.
Americans clearly want government regulation on AI
At this point, even the AI companies are asking for government regulation. OpenAI put out a statement on Sept. 9 asking to work with Congress on federal guidelines for artificial intelligence. Clearly, these companies are unable to set these guidelines themselves, and need the federal government to step in to tell them where the line in the sand is.
Steven Kull, founder of the Program for Public Consultation at the University of Maryland’s School of Public Policy, wrote recently for USA TODAY Opinion that both sides of the political aisle are looking for stronger AI regulation from the government. According to the survey his group conducted, 85% of Democrats and 79% of Republicans want a federal agency to oversee AI, while 82% of Democrats and 78% of Republicans want some form of government testing for AI systems.
Though the American people clearly want this, there is little movement in Congress toward a system for regulation. House Speaker Mike Johnson, R-Louisiana, said on Sept. 14 that he called for a meeting with AI executives, but he said the day prior that he would not pass legislation that could potentially put the United States at risk of falling behind China’s AI development.
Both Republicans and Democrats within Congress have pushed for AI regulations in recent months with minimal success. The problem is that they aren’t working together, nor are they working with the companies that are creating artificial intelligence to better understand the way things are going, and the people setting the agenda in our government have no desire to do anything that might affect the country’s bottom line.
They must get over that fear of losing to China if we are to keep the nation from torpedoing toward a doomsday scenario.
This is a moment where bipartisanship could prevail and politicians could offer wins to both sides of the aisle. People are frightened of the future of AI, and they want to see regulation. While I doubt Congress would pass a moratorium on AI development, as suggested in the bill Sen. Bernie Sanders, I-Vermont, filed recently, I think that there are some things that both political parties could agree on, like a nationwide data center moratorium or regulations on when and how AI is used. At this point, they have no choice – the fate of humanity might very well be in their hands.
Follow USA TODAY columnist Sara Pequeño on Bluesky: @sarapequeno.bsky.social