Anthropic CEO Dario Amodei recently pushed back against the idea that he\u2019s been painting an overly pessimistic picture of artificial intelligence and how it might shape the future.<\/p>\n
Amodei\u2019s comments came in response to investor Gavin Baker, who argued \u2014 both on the All-In podcast<\/a> and on X<\/a> \u2014 that Amodei\u2019s warnings about the dangers of AI have helped to fuel a backlash<\/a> in the United States, particularly against data centers<\/a>.\u00a0<\/p>\n Claiming that Amodei has \u201clost the argument\u201d when it comes to AI regulation (Anthropic has advocated for some regulations, including a California bill that imposes transparency requirements on large AI companies<\/a>), and given that \u201che is about to be the CEO of one of the most important companies in the world,\u201d Baker wrote, \u201cI respectfully think he should make an effort to be a more positive advocate for his own industry.\u201d<\/p>\n Baker is far from the only one arguing that AI skepticism and even government crackdowns are a natural response to the dire warnings of some AI executives<\/a>. But in a series of posts<\/a>, Amodei disagreed with the idea that his \u201cmessaging has been disproportionately negative.\u201d Instead, he said that his writing has been \u201cabout equally balanced between risks and benefits,\u201d and that he wrote his essay \u201cMachines of Loving Grace\u201d<\/a> because he \u201cdidn\u2019t feel the AI industry was painting an inspiring enough picture of how the technology could radically transform the world for the better.\u201d<\/p>\n Nonetheless, Amodei acknowledged that \u201cthe public has a negative view of AI\u201d and he agreed that \u201cthis is a big problem.\u201d Where he disagreed was with the idea that this negativity is \u201cprimarily caused\u201d by Amodei \u201cor any other AI leader warning about AI\u2019s risks.\u201d<\/p>\n \u201cI think it is fundamentally a crisis of trust,\u201d Amodei said. \u201cI think that ordinary people don\u2019t trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over.\u201d<\/p>\n Indeed, \u201ctrust\u201d is a word that often comes up in headlines about OpenAI CEO Sam Altman<\/a>, and is apparently an issue for other AI CEOs like Amodei, too<\/a>. In Amodei\u2019s telling, however, this is a crisis that\u2019s been decades in the making<\/a>, with the AI backlash \u201cjust the latest iteration of it.\u201d<\/p>\n \u201cI think by far the most accurate criticism of AI companies including Anthropic is that we haven\u2019t yet delivered on our big promises to benefit the world,\u201d Amodei said. \u201cThat is totally on us, and I think it\u2019s the criticism you should be making, instead of all this stuff about messaging and marketing.\u201d<\/p>\n Put another way, he said promising that AI will cure cancer is \u201cmore a cliche than it is inspiring\u201d; what would actually change people\u2019s opinions of AI would be \u201cactually curing cancer.\u201d<\/p>\n As for regulation<\/a>, Amodei argued that Baker was painting \u201ca false choice\u201d between distributing AI widely without regulation, or concentrating the technology in the hands of a few companies through regulation.<\/p>\n \u201cI know that there\u2019s a sort of Silicon Valley shorthand where regulation = regulatory capture = concentration of power, but I\u2019ve always found this to be an overly simplified picture of the world,\u201d Amodei said. \u201cMany people outside this bubble think of regulation as something that constrains corporate power and benefits ordinary people.\u201d<\/p>\n Amodei added that he doesn\u2019t \u201cnecessarily agree with that perspective either,\u201d but he said that\u2019s \u201cwhy Anthropic has always made its policy proposals very carefully.\u201d<\/p>\n \u201cWe try very hard to make proposals that disadvantage (slow down) frontier AI companies while *advantaging* smaller competitors,\u201d he said. <\/p>\n Why make those proposals at all? Amodei said that in his view, \u201cAI is *structurally* a technology that tends to concentrate power\u201d and that \u201copen-weights do help some with this but are nowhere near a sufficient solution because they simply shift the concentration somewhat to those with the most compute and chip.\u201d<\/p>\n \u201cBy contrast I think the right \u2018rules of the road\u2019 can simultaneously (a) address AI\u2019s cyber\/bio\/alignment risks, (b) institutionally constrain the power of the frontier AI companies, and (c) leave room for open-weights models while also addressing the specific risks that they bring,\u201d he said.<\/p>\n This post has been updated to include more quotes from Amodei\u2019s posts, including those around open weights.<\/p>\n 2\/2 Second, on the messaging around AI.\u00a0 I do not agree that my messaging has been disproportionately negative.\u00a0 In fact it has been about equally balanced between risks and benefits: I\u2019ve written one major essay about each, and even in interviews where I discuss the risks, I\u2026<\/p>\n \u2014 Dario Amodei (@DarioAmodei) August 15, 2026<\/a><\/p>\n