On July 9, 2026, Anthropic — which identifies itself as a Public Benefit Corporation — launched a public initiative inviting people to send in their toughest questions about artificial intelligence. The company says it will publicly track and report the concrete steps it takes to address those questions and will be transparent about where it falls short of its stated goals.
Why they launched this
Anthropic frames the effort as a response to mixed public sentiment about AI. Many people already use AI daily and see its potential to reduce labor, change how we learn, accelerate scientific and technological progress, create new sources of prosperity, and help solve major medical and social problems. At the same time, significant concerns persist: potential job losses, the devaluation of creative work, effects on human agency and relationships, and the risks if powerful capabilities end up in the wrong hands.
The company argues that acting in the public interest requires understanding these hopes and fears, so it is expanding its research and engagement work to capture a broader range of public views.
What Anthropic has already done
Anthropic outlines prior research and engagement that form the basis for the new call for questions:
- Anthropic Public Record: in its first round, the survey asked 52,000 Americans to describe their biggest hopes and concerns about AI.
- Anthropic Interviewer: through this tool, Anthropic surveyed about 81,000 Claude users across 159 countries and in 70 languages.
- In-person focus groups and convenings: the company held dozens of in-person focus groups and met with communities whose work and traditions are especially relevant to AI-related questions.
- Study of anonymized real-world usage data: Anthropic has been analyzing anonymous usage of Claude.
Governance, programs, and safeguards
Anthropic also lists organizational structures and programs intended to advance its public-benefit mission:
- Anthropic Institute: an internal research effort focused on addressing the most significant social challenges posed by AI.
- Long-Term Benefit Trust: a governance mechanism, in place since early in the company’s history, that provides impartial oversight of progress toward the public-benefit mission.
- Investments in safeguards and research into model behavior and internal workings to help align models with beneficial goals.
- Free provision of models to scientists to support research and medicine.
- A fellowship program that pairs early-career Claude users with nonprofits to extend AI’s benefits.
What they are asking and promising
Anthropic asks the public to raise hard questions about AI’s impacts on jobs, society, and families; about harnessing AI’s potential for science and medicine; and about the broader trajectories for one of the most powerful technologies in human history.
In return, Anthropic pledges to:
- publicly track and report specific actions taken to address submitted questions;
- be explicit about where it does not meet its goals or where limitations exist.
People can view questions others have asked and submit their own via Anthropic’s “hard questions” website.
Why this matters
The initiative represents an example of a major AI company promising a more open dialogue and public accountability. The company’s prior surveys—52,000 Americans and 81,000 users from 159 countries—along with its internal research entities and oversight mechanism indicate an attempt to ground policy and development decisions in empirical engagement. At the same time, delivering on the transparency and follow-through Anthropic promises will require ongoing public scrutiny and verification of the company’s reported progress.
