According to Business Insider, Anthropic CEO Dario Amodei is well known for frequently warning about the possible risks of his own AI products, but this has also cost his company.In an article published this month, Amodei warned that the power of AI “has become unquestionable.”. As an example, he cited Anthropic's latest model, Mythos, as posing "very real risks" to cybersecurity, the financial industry, critical infrastructure and national security.
Amodei
He then called on the U.S. government to take stronger intervention measures to deal with these risks. But on Friday, his warning seemed to backfire: The U.S. government suddenly imposed export controls on Anthropic's latest model.
Anthropic said it has cut off access to its Mythos 5 and Fable 5 after the Trump administration ordered it to block foreign users from accessing the models. The Pentagon's chief information officer expressed support for the move in a post on the X platform, writing: "Some things are more important than revenue cycles, hype, and pre-IPO valuations."
This sudden development has sent the entire technology industry into panic. Gary Marcus, an AI researcher and well-known skeptic, called the Trump administration’s move “a gross overreaction and counterproductive.”
Yang Likun said Amodei was reaping the consequences
However, "AI Godfather" Yann LeCunn placed the blame on Amodei. "Amodei has been spreading ridiculous alarmism about the Mythos/Fable model (and the entire AI field), and now it has finally paid off," Yang Likun said, "You will reap what you sow."
Long warning about AI risks
Amodei has long tried to portray himself as the “sane adult” in the AI industry.He was an integral researcher at OpenAI until he left the company and founded Anthropic out of concern that OpenAI CEO Sam Altman was prioritizing the speed of product releases rather than ensuring their safety.
Since then, he has issued a series of statements, saying that what he and OpenAI, Google, Meta, Amazon and other companies are developing and releasing to the public may fundamentally change the way of life that humans are familiar with. This impact can be both positive and negative.
Amodei has warned that AI could eliminate half of entry-level white-collar jobs and cause unemployment to soar to levels not seen since the COVID-19 pandemic, possibly even recreating the unemployment shock of the 2008 global financial crisis (Great Recession).
While Amodei has softened in recent weeks on the impact of AI on jobs as Anthropic prepares to go public, he and Anthropic have remained unreserved in their warnings about other security risks.
“Cybersecurity risks posed by Mythos-level models will not be the last risks we have to face,” he said in an article published in June. “I think biosecurity risks may soon follow, and more serious AI autonomy risks will not be too far away.”
He also said that legislators’ current level of concern and awareness about AI can no longer keep up with the rapid development of AI technology.
Earlier this month, Anthropic called for a moratorium on the development of cutting-edge AI models in a paper, warning that the latest models were getting closer to self-improvement, which "may increase the risk of humans losing control of AI systems."
"We believe it would be a good thing if the world could choose to slow down or temporarily pause the development of cutting-edge AI so that social structure and alignment research can keep up with technological advances." Anthropic said.
Ironically, this "pause button" was finally pressed last Friday, but the target became Anthropic itself.