People have many difficult questions about AI - questions like who sets the rules, whether it can improve their children's futures, whether it makes the world more dangerous, and whether it can help cure diseases. Anthropic sees it as the company's responsibility to address them.
Many people view AI favorably. They already use it daily and recognize its potential to reduce tedious work, transform learning, accelerate scientific and technological progress, create new prosperity, and tackle major medical and social challenges.
However, many also harbor serious concerns. Some worry about AI contributing to job losses. Others fear it could diminish the value of creative work. Questions about human agency arise too - whether AI might undermine people's ability to think independently, form human connections, and find meaning in their lives. Many wonder what happens when AI's capabilities end up in the wrong hands and whether the benefits truly outweigh the costs.
As a Public Benefit Corporation, Anthropic's mission is to ensure that advanced AI models deliver benefits while mitigating risks. That mission has driven the company to invest in AI safeguards to reduce misuse, to research the behavior and inner workings of AI models for alignment with beneficial goals, to provide AI models at no cost to scientists, and to launch a fellowship program that connects early-career Claude users with nonprofits to broaden AI's benefits.
To genuinely serve the public interest, Anthropic needs to fully understand the hopes and concerns people have about this technology. As part of a new initiative focused on hard questions, the company is taking additional steps to learn people's views on various AI-related topics and transparently track progress toward its public benefit goals.
Groundwork has already been laid through surveys of thousands of people about their views on AI:
- The Anthropic Public Record - a public survey that in its first round asked 52,000 Americans to share their biggest hopes and concerns about AI;
- A survey of 81,000 Claude users across 159 countries and 70 languages using the Anthropic Interviewer;
- Dozens of in-person focus groups and convened sessions with groups whose work and traditions are relevant to questions raised by AI; and
- Studies of Claude usage through anonymized, real-world data.
Anthropic created the Anthropic Institute - a research effort within the company aimed at confronting the most significant challenges AI will pose to society. Since early in the company's history, the Long-Term Benefit Trust has provided impartial oversight on how effectively Anthropic advances its public benefit mission.
Now, Anthropic is explicitly inviting the public to submit their hardest questions about AI: questions about AI's effects on jobs, society, and families; questions about realizing AI's dramatic potential for science and medicine; and questions about where one of the most powerful technologies in human history might lead next.
In return, Anthropic will publicly track and report the specific actions being taken to address those questions - and will be transparent about any ways in which the company may fall short of its stated goals.
To see questions others are asking and to share your own, visit the hard questions website.