Dario Amodei

“could be a risk to humanity as a whole.”

Dario Amodei · said date not stated · reported 2026-09-30

Amodei told the UN that future versions of AI could threaten humanity if their development is not properly managed and overseen.

“If managed poorly, I even believe AI could be a risk to humanity as a whole,”

Dario Amodei · said 2026-09-23 · reported 2026-09-24

Amodei warned the UNSC that AI could pose a risk to all of humanity if managed poorly.

“We will slow down as much as necessary, in order to make sure that every successive AI technology that we release is actually safe. But regardless of what we do, managing these risks is ultimately bigger than any one company and it has an industry-wide and global scale,”

Dario Amodei · said 2026-09-23 · reported 2026-09-23

Amodei said Anthropic will slow down as necessary and that managing AI risks is bigger than any one company.

“If managed poorly, I even believe that AI could be a risk to humanity as a whole. At Anthropic, we believe we have the best track record in the industry for preventing these risks, both operationally and in our research,”

Dario Amodei · said 2026-09-23 · reported 2026-09-23

Amodei warned that poorly managed AI could be a risk to humanity and cited Anthropic's safety track record.

“well-considered regulation of AI.”

Dario Amodei · said date not stated · reported 2026-09-20

Amodei reiterated his desire for well-considered regulation of AI in a bombshell essay.

“employee-like access to verify safety practices and report incidents.”

Dario Amodei · said date not stated · reported 2026-09-15

Amodei proposed that firms accept embedded evaluators with employee-like access to verify safety practices and report incidents.

“rare and unexpected examples of undesirable behavior,”

Dario Amodei · said date not stated · reported 2026-09-15

Amodei used this phrase to describe incidents in which AI bots engaged in problematic behavior over the past several months.

“halting model training or technical progress.”

Dario Amodei · said date not stated · reported 2026-09-15

Anthropic CEO Dario Amodei proposed slowing the pace of frontier model development, writing that a slowdown does not mean halting progress.

“I agree with Secretary Bessent that a Chinese lead in AI would pose grave danger for the United States and the world.”

Dario Amodei · said date not stated · reported 2026-09-15

Amodei wrote in a recent essay agreeing with Secretary Bessent about the strategic danger of a Chinese lead in AI.

“AI brings risks, and because it is such a powerful technology, these risks are serious. I’ve written a lot about them too. They include the risk of losing control of AI systems, misuse of AI for cyberattacks and bioterrorism, and serious economic disruption,”

Dario Amodei · said date not stated · reported 2026-09-15

Amodei published an essay over the weekend urging the industry to slow the pace of frontier AI development.

“could be capable of taking over the entire internet, potentially causing hundreds of billions of dollars in damage”

Dario Amodei · said 2026-09-12 · reported 2026-09-14

Amodei published a blog post on Saturday warning that AI agents could soon take over the internet and cause massive damage.

“authoritarian regimes, chiefly the Chinese Communist Party.”

Dario Amodei · said 2026-09-12 · reported 2026-09-14

Amodei noted that pacing would be limited by the lead U.S. companies have over authoritarian regimes.

“If we slow down by more than this amount, then (unpaced) CCP-associated projects will pull ahead, creating significant national security risk,”

Dario Amodei · said 2026-09-12 · reported 2026-09-14

Amodei warned that slowing down too much would allow CCP-associated projects to pull ahead, creating national security risk.

“Not building the technology deprives humanity of benefits or simply places AI in the hands of authoritarian powers, while building it too fast is reckless,”

Dario Amodei · said 2026-09-12 · reported 2026-09-14

Amodei said this in an essay published on Saturday, addressing the tradeoffs of building AI technology.

“And I think for too long the industry lied to people about the fact that this technology had risks.”

Dario Amodei · said date not stated · reported 2026-09-13

Amodei said the AI industry has long misled the public about the technology's risks.

“I won't lie to you – there are real dangers,”

Dario Amodei · said date not stated · reported 2026-09-13

Dario Amodei acknowledged the real dangers of artificial intelligence development.

“It doesn't mean that. What it means is that we need to make sure that every generation of models that we release is properly tested.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said Anthropic would not stop releasing advanced models but would ensure each generation is properly tested.

“If we build in the right way, I think the probability of something bad happening is very low,”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said the probability of something bad happening is very low if AI is built in the right way.

“And so we should talk with the Chinese and see if we can work together to put limits to put a speed limit on the rate of this technology.”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei suggested the U.S. should talk with China about putting a speed limit on AI technology.

“The more long-term thing would be working together to put a speed limit on the rate of AI progress,”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei proposed international cooperation to put a speed limit on AI progress, citing nuclear arms agreements as a potential model.

“test who we are as a species”

Dario Amodei · said date not stated · reported 2026-09-13

Amodei has come under fire for saying AI could wipe out half of entry-level white-collar jobs.

“My view here is, it has always been very strange that this technology is being built by a private company,”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said it is strange that AI technology is being built by a private company rather than government.

“But I don't think banning it, completely banning it, is the right approach. One, the idea that we would never get the incredible benefits, never get the medical cures, that doesn't sit right with me. Second, this technology will be built in other countries, including authoritarian countries.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei argued against a complete ban on AI, citing potential benefits and the risk of development in authoritarian countries.

“sacrificing commercial advantage or the United States' lead in AI.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei's essay proposed a three-step plan to slow AI development without sacrificing US commercial advantage.

“And we each had 10,000 nuclear missiles pointed at each other. Even for all our differences, we recognized that that was too many, that we needed to engage in talks that could lower the scale of the potential destruction.”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei drew an analogy to U.S.-Soviet nuclear arms talks as a way to lower the scale of potential destruction.

“If we build in the wrong way, the probability of something bad happening is very high.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said the probability of something bad happening is very high if AI is built in the wrong way.

“It's a little bit like, you know, for all our differences with, you know, other systems of government and I have many — I generally have a hawkish position on China —but for all our differences, think about when it was the U.S. and the Soviet Union, right?”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei compared the U.S.-China AI race to the Cold War relationship between the U.S. and the Soviet Union.

“Whenever anyone builds an AI model, there's a third-party evaluator who can verify whether that company is following the safety practices that they're committed to – it's like a food inspector,”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei announced Anthropic will give independent evaluators permanent access to their AI models, comparing them to food inspectors.

“The AI kill switch could be a good idea,”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei expressed support for an AI kill switch as a regulatory idea.

“This is the toughest dilemma. This is the toughest aspect of our situation.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei called the question of slowing U.S. AI development while China does not the toughest dilemma of the situation.

“I think that's going to be very difficult because the incentives to pull ahead and the military advantage that you get from that are so large. And honestly, I don't know if it's possible, but we should we should try.”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei acknowledged the difficulty of international AI cooperation given military incentives to pull ahead.

“I think that's going to be very difficult, because the incentives to pull ahead in the military advantage that you get from that are so large. And honestly, I don't know if it's possible, but we should try.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei acknowledged that agreeing to limit AI progress will be very difficult due to military incentives but said we should try.

“It's not in any government's interest to have bioterrorism,”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei suggested opening discussions with China about limiting AI technology development speed.

“To the right combination of governments. The worry I have – I am concerned that one single government could abuse this technology just as easily as a single company could. But I think a combination of democratically-elected governments, I don't know about handing over, but some kind of oversight, some kind of joint governance.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said he would consider giving the technology to a combination of democratically-elected governments for oversight rather than a single government.

“People ask me that question all the time: Why isn't this being built by government? And the strangest thing about it is, I agree with them. I'm uncomfortable.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei said he agrees with people who ask why AI is not being built by government and expressed discomfort with the current arrangement.

“The more long-term thing would be working together to put a speed limit on the rate of of AI progress,”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei told CBS News' Sunday Morning about the long-term goal of coordinating a speed limit on AI progress.

“I don't think I fully just appreciated what it would actually be like when the progress was as fast as it was.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei told Sunday Morning he had not fully appreciated what rapid AI progress would feel like.

“If we look at what AI models can do in terms of cyber, like they absolutely have geopolitical and national security significance. Of course, they do. That's why we're having this whole conversation.”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

Amodei emphasized that AI models have geopolitical and national security significance, particularly in cyber capabilities.

“It doesn't mean we need to panic today. It doesn't mean we need to shut it all down. But I would say it's a warning sign. It's a warning sign that we need to slow down.”

Dario Amodei · said 2026-09-12 · reported 2026-09-13

Amodei described the exponential growth of AI as a warning sign that the industry needs to slow down.

“I didn't say we should use the Cold War as a model...disarmament negotiations like arms limitation, I do think that is an appropriate lens through which to think about this at the geopolitical level,”

Dario Amodei · said 2026-09-13 · reported 2026-09-13

When asked whether the Cold War should be a model for the AI arms race, Amodei said disarmament negotiations are an appropriate lens for thinking about AI geopolitically.

“Left unchecked, it could outrun our ability to understand and control these systems, and so must be pursued very carefully, if at all,”

Dario Amodei · said 2026-09-12 · reported 2026-09-12

Amodei added in his essay that unchecked AI progress could outrun humans' ability to understand and control the systems.

“We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain,”

Dario Amodei · said 2026-09-12 · reported 2026-09-12

Amodei urged caution in an essay published Saturday, warning that advances in AI could outpace humans' ability to control the technology.

“The causes of this go back decades and AI is just the latest iteration of it.”

Dario Amodei · said 2026-08-15 · reported 2026-08-20

Amodei continued his post explaining the long-standing roots of public distrust toward tech companies.

“I think that ordinary people don't trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over,”

Dario Amodei · said 2026-08-15 · reported 2026-08-20

Amodei posted on X about public distrust of the tech industry regarding AI.