Public transcript

The Co-Founders of Claude AI Tell Oprah About the Impact Artificial Intelligence Has on Your Life

By: Oprah
Length: 1 h 6 min
Language: English
Text from: the video's captions
Transcribed: 13 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

Oprah interviews Anthropic co-founders and siblings Dario and Daniela Amodei about the promise and dangers of artificial intelligence and their chatbot Claude. They discuss why they left OpenAI, their public benefit corporation structure, their 80% wealth pledge, and decisions like banning users under 18, refusing ads, and rejecting military uses for autonomous weapons and domestic mass surveillance. They also cover the new Mythos model's cyber capabilities, job disruption, and why the public should learn about AI rather than fear it. Their central message is that AI is an epochal change that must involve everyone, guided by trust and principle.

Key points

Dario Amodei compares AI to epochal shifts like fire and the Industrial Revolution, saying you cannot stop the train but you can steer it.
Anthropic is incorporated as a public benefit corporation, and all seven co-founders have pledged 80% of their wealth to philanthropic causes.
Anthropic does not allow users under 18 on Claude and uses interaction patterns to detect and suspend accounts of suspected minors.
The company refuses advertising, relying on subscriptions and enterprise sales so the incentive is usefulness rather than maximizing user engagement.
Anthropic refused to let the military remove guardrails for fully autonomous weapons and domestic mass surveillance, risking the company's survival.
Dario wrote a New York Times op-ed against a bill that would have banned all AI regulation; the bill was voted down 99 to 1 in the Senate.
The new model Mythos proved unusually good at finding exploitable code, so Anthropic gave it to about 40 defender companies before wider release.
Daniela notes women use AI far less than men and urges the public to learn the technology so they can use their voice effectively.
Dario says he does not always sleep well, and his biggest worry is whether they will get the big things right despite good intentions.
Both founders say a well-lived life means agency, dignity, principle, and being the best version of yourself, which AI could either damage or enable.

Questions it answers

00:00Opening and Claude's question

How does Dario justify building AI he has said could cause human extinction?

Dario compares AI to epochal changes like fire and the Industrial Revolution, arguing the technology is happening regardless and that Anthropic's role is to steer the train and decrease the risk, hopefully to zero.

Oprah asked Claude itself to generate the hardest interview question.
Dario says you cannot stop the train but you can steer it away from the rocks.
He acknowledges commercial applications and funding needs are entangled with safety goals.

05:47Founders and company mission

Who are the Amodei siblings and why did they start Anthropic?

Dario and Daniela are siblings who always wanted to do good in the world, with Dario in biophysics and Daniela in global development. They left OpenAI to build a company committed to making AI go well for everybody, structured as a public benefit corporation with an 80% wealth pledge.

They preserve sibling time weekly, playing video games and watching movies.
Anthropic is a public benefit corporation balancing commerce and public benefit.
All seven co-founders pledged 80% of their wealth to philanthropic causes.

14:07Benefits versus harms

Why does the negative AI story dominate right now?

Dario explains that harms like AI-influenced suicides happen quickly while benefits like curing cancer take years of research and clinical trials. Anthropic bought a small biotech company to speed up AI-driven drug design for cancer, Alzheimer's, and heart disease.

Harms are visible now; benefits like new drugs take years.
Daniela imagines her children growing up with entire classes of diseases simply cured.
Dario says the real goal is helping people, not fixing the narrative.

18:48Attachment, kids, and incentives

How does Anthropic protect children and prevent unhealthy attachment to Claude?

Anthropic bans users under 18 and uses interaction patterns to detect minors, suspending accounts until age verification. Because Anthropic has no ads and relies on subscriptions, its incentive is to be useful rather than to keep users addicted or engaged.

Dario calls falling in love with AI a bad idea and a real danger that is already happening.
Kids sometimes give themselves away by asking about arthritis medication first.
The no-ads model means Claude is designed to end conversations once you have your answer.

28:58Standing up to the Pentagon

What happened when the military wanted safety guardrails removed from Claude?

Anthropic refused to allow fully autonomous weapons and domestic mass surveillance, even though the Pentagon implied it could stop other companies from doing business with them. The co-founders unanimously accepted the risk that it could end the company.

Anthropic had been the first AI company to work with national security.
The founders asked whether their past selves would be proud if they compromised.
Daniela frames it as a policy disagreement, not a political one.

36:36Regulation and public voice

What should AI regulation look like and what can the public do?

Dario describes speaking out against a bill that would have banned all AI regulation, which was voted down 99 to 1, and Anthropic's support for California's SB 1047. Daniela urges people to become literate in AI so they can form opinions and use their voice with companies and government.

Dario says trust is in short supply and AI can only diffuse at the speed of trust.
Daniela notes women use AI far less than men and encourages engagement.
Claude has a Constitution that citizens have given feedback on.

44:38Jobs and human qualities

Will AI wipe out jobs and make humans unnecessary?

Dario has said 50% of entry-level jobs could be wiped out if the technology is deployed without helping people adapt. Daniela argues jobs will change rather than vanish: AI may diagnose illness while doctors provide bedside manner, physical examination, and human relationship.

A better world is possible but will not happen on its own.
The qualities we prize in workers will shift toward human connection.
Patients have better clinical outcomes with doctors they have a personal relationship with.

49:47The Mythos model

What is the new Mythos model and why is it dangerous?

Mythos is Anthropic's latest model with a particularly large capability jump. It turned out to be much better than human software engineers at finding exploitable code, making it effective at both conducting and defending against cyber attacks, so Anthropic gave it to about 40 defender companies before wider release.

One company fixed more bugs with Mythos in a week than in the previous year.
The hope is a more secure world with less ransomware by the end of the year.
There is a dangerous transition period while defenders patch vulnerabilities.

Notes

Overview of the Conversation

Oprah interviews Dario and Daniela Amodei, the sibling co-founders of Anthropic, the company behind the AI assistant Claude, reportedly valued at $900 billion.
Oprah opens by asking Claude itself for the toughest question: how Dario justifies racing to build a technology he has said could cause human extinction.
Dario frames AI as an epochal change like fire or the Industrial Revolution: "You can't stop the train. But what you can do is you can steer the train."

Company Values and Structure

Anthropic is incorporated as a public benefit corporation, legally balancing commercial interests with public benefit.
All seven co-founders have pledged 80% of their Anthropic wealth to philanthropy and are thinking about starting a foundation.
The company refuses advertising, relying on subscriptions and enterprise sales, so its incentive is usefulness rather than maximizing hours of engagement.
Claude is not available to users under 18; interaction patterns help detect minors, and suspicious accounts are suspended pending age verification.

Hard Decisions

Anthropic refused military use cases involving fully autonomous weapons and domestic mass surveillance, even though the Pentagon suggested it could block other companies from doing business with them.
The co-founders agreed unanimously, accepting the risk it could end the company.
Dario wrote a New York Times op-ed against a bill that would have banned all AI regulation; it was voted down 99 to 1 in the Senate.
Anthropic publicly supported California's SB 1047 despite industry opposition and imperfect drafting.

Benefits, Risks, and Jobs

Benefits like curing cancer take years, while harms happen quickly, which is why negative narratives dominate.
Dario has said 50% of entry-level jobs could be wiped out if the technology is deployed without helping people adapt.
Daniela argues jobs like doctors will change rather than vanish: AI may handle diagnosis while humans provide bedside manner and touch.

The Mythos Model

Mythos showed a large jump in capability and proved far better than human engineers at finding exploitable code.
Anthropic gave it to roughly 40 defender companies, including major banks and software providers, before wider release, hoping to end the era of ransomware.

Closing Messages

Dario urges the public not to be afraid but to learn, lean in, and be critical.
Daniela emphasizes "light and shade": both the vast benefits and the real risks are true at once.
Both define a well-lived life through agency, dignity, principle, and being the best version of yourself.

Text from the video's own captions. Summary and notes written by AI from the transcript, so check anything important against the recording.