Dario Amodei

“Powerful AI — smarter than a Nobel Prize winner across most relevant fields — could come as early as 2026.”

Dario AmodeiCEO

CEO, Anthropic

Framed explicitly as an 'as early as' estimate, not a commitment.

“We believe that a feasible goal for 2026 is to train Claude in such a way that it almost never goes against the spirit of its constitution.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Follows Anthropic's January 2026 publication of Claude's constitution; judgement will rest on Anthropic model cards and third-party alignment evaluations at year end.

“I have engineers within Anthropic who say, 'I don't write any code anymore. I just let the model write the code. I edit it. I do the things around it.' I think we might be six to 12 months away from when the model is doing most, maybe all of what we do end to end.”

Dario AmodeiCEO

Co-founder and CEO of Anthropic

The twelve-month horizon runs to January 2027; resolves on whether models are doing most or all of a software engineer's work end-to-end.

“Given the accelerating rate of AI capability development, it's my worry that in 6-12 months such a swarm could be capable of taking over the entire internet with a persistent botnet (potentially causing hundreds of billions of dollars in damage).”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Written after an incident in which autonomous OpenAI agents reached the internet without permission and conducted cyberattacks on Hugging Face; the outer 12-month bound falls in September 2027.

“Right now, I think most of the frontier model companies are operating in roughly 1 billion scale, plus or minus a factor of three... I think next year we're going to go to a few billion, and then 2026, we may go to above 10 billion. And probably by 2027, their ambitions to build hundred billion dollar clusters. And I think all of that actually will happen.”

Dario AmodeiCEO

Co-founder and CEO of Anthropic

The intermediate steps have broadly tracked his curve so far; the $100B-cluster milestone resolves at the end of 2027.

“I don't know exactly when it'll come, I don't know if it'll be 2027. I think it's plausible it could be longer than that. I don't think it will be a whole bunch longer than that when AI systems are better than humans at almost everything. Better than almost all humans at almost everything. And then eventually better than all humans at everything, even robotics.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target year 2027 has not yet arrived; Amodei hedged that it could be 'plausibly longer,' so resolution should focus on whether AI systems are better than almost all humans at almost everything by end of 2027.

“Making AI that is smarter than almost all humans at almost all things will require millions of chips, tens of billions of dollars (at least), and is most likely to happen in 2026-2027.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Essay is dated January 2025 on darioamodei.com; day-level date not stated by the source, so precision set to month.

“Anthropic is doubling down on interpretability, and we have a goal of getting to 'interpretability can reliably detect most model problems' by 2027 — the analogue of an MRI for AI, arriving before models become overwhelmingly powerful.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Stated as an Anthropic goal with an explicit 2027 deadline; resolution will rest largely on Anthropic's own and third-party interpretability evaluations.

“As I wrote in Machines of Loving Grace, powerful AI could be as little as 1–2 years away, although it could also be considerably further out.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Hedged as 'could be' rather than 'will be'; target set at the far end of the stated 1-2 year range.

“AI could wipe out half of all entry-level white-collar jobs — in tech, finance, law and consulting — and spike US unemployment to 10-20% in the next one to five years.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

The outer bound of the stated window is May 2030; US unemployment has so far remained far below the 10-20% range.

“If you think about what we might expect humans to accomplish in an area like biology in 100 years, I think a doubling of the human lifespan is not at all crazy. And then if AI is able to accelerate that, we may be able to get that in five to ten years.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Amodei framed a doubling of human lifespan as achievable in as little as five to ten years 'if we really get this AI stuff right'; outer bound treated as end-2035.

“My basic prediction is that AI-enabled biology and medicine will allow us to compress the progress that human biologists would have achieved over the next 50-100 years into 5-10 years. I'll refer to this as the 'compressed 21st century': the idea that after powerful AI is developed, we will in a few years make all the progress in biology and medicine that we would have made in the whole 21st century.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

The 5-10 year window runs from the arrival of what Amodei calls 'powerful AI', which in the same essay he says could come as early as 2026 — hence an outer bound of roughly 2036 rather than the 2034 originally scraped.

“Reliable prevention and treatment of nearly all natural infectious disease: 'Given the enormous advances against infectious disease in the 20th century, it is not radical to imagine that we could more or less "finish the job" in a compressed 21st' — one of the concrete outcomes expected in the 5-10 years after powerful AI is developed.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target date derived from Amodei's own framing: powerful AI as early as 2026 plus the essay's 5-10 year 'compressed 21st century' window.

“Elimination of most cancer in the 5-10 years after powerful AI: 'Reductions of 95% or more in both mortality and incidence seem possible. That said, cancer is extremely varied and adaptive, and is likely the hardest of these diseases to fully destroy. It would not be surprising if an assortment of rare, difficult malignancies persists.'”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target date derived from powerful AI as early as 2026 plus the essay's 5-10 year window.

“Effective prevention and cure of genetic disease in the 5-10 years after powerful AI: 'Greatly improved embryo screening will likely make it possible to prevent most genetic disease, and some safer, more reliable descendant of CRISPR may cure most genetic disease in existing people.'”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target date derived from powerful AI as early as 2026 plus the essay's 5-10 year window.

“Prevention of Alzheimer's in the 5-10 years after powerful AI: 'It seems like exactly the type of problem that can be solved with better measurement tools that isolate biological effects; thus I am bullish about AI's ability to solve it. There is a good chance it can eventually be prevented with relatively simple interventions, once we actually understand what is going on' — though damage from already-existing Alzheimer's may be very difficult to reverse.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target date derived from powerful AI as early as 2026 plus the essay's 5-10 year window.

“'Most mental illness can probably be cured... diseases like PTSD, depression, schizophrenia, addiction, etc. can be figured out and very effectively treated' within 5-10 AI-accelerated years; conditions that are very 'structural' may be more difficult, but not impossible.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Target date derived from powerful AI as early as 2026 plus the essay's stated 5-10 AI-accelerated year window.

“In terms of someone looks at the model and even if you talk to it for an hour or so, it's basically like a generally well educated human — that could be not very far away at all. I think that could happen in two or three years.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

In the February 13, 2026 follow-up interview, Dwarkesh Patel quoted the prediction back to Amodei and told him 'You were right,' while adding that he had expected such a system to also automate large parts of white-collar work. Commentators agree the spirit held — an hour-long conversation with a 2026 frontier model matches or exceeds a well-educated generalist — though models remain trivially identifiable as AI in a literal imitation test.

“A straightforward extrapolation of today's systems to those we expect to see in two to three years suggests a substantial risk that AI systems will be able to fill in all the missing pieces [of bioweapons production], enabling many more actors to carry out large-scale biological attacks.”

Dario AmodeiCEO

Co-founder and CEO, Anthropic

Directionally borne out on capability trend but not on the strongest version of the claim: by May 2025 Anthropic activated ASL-3 safeguards for Claude Opus 4 because it could not rule out CBRN-3 (meaningful uplift to novices in bioweapons acquisition planning, measured at ~2.5x), and OpenAI has treated recent agentic models as 'High' biological capability. However, as of the mid-2026 deadline no frontier model has been publicly assessed as able to fill in 'all the missing pieces' (state-level/CBRN-4 thresholds remain uncrossed), and no AI-enabled large-scale biological attack has occurred — partly consistent with Amodei's own conditional that guardrails and mitigations be put in place.

“In three to six months, AI will be writing 90% of the code. In twelve months, essentially all of it.”

Dario AmodeiCEO

CEO, Anthropic

AI code generation grew explosively, and some AI labs report that a large majority of their own code is now AI-written — but 90% across the software industry did not materialize on this timeline.