Anthropic asks the public to submit its hardest AI questions

Anthropic is opening a new public initiative focused on the questions people find most difficult—and consequential—about artificial intelligence.

The company is asking people to submit concerns and ideas spanning AI’s effect on jobs, creative work, family life, human agency, public safety, medicine, and scientific progress. Anthropic says it plans to publish the actions it takes in response and report openly where it fails to meet its public-benefit goals.

Building on public input

The announcement points to several existing research efforts behind the initiative. Anthropic says it has surveyed 52,000 Americans through the Anthropic Public Record, interviewed 81,000 Claude users in 159 countries and 70 languages, held focus groups and community sessions, and analyzed anonymized real-world Claude usage.

The company also highlights its Anthropic Institute, safety research, AI-for-science programs, nonprofit fellowship work, and Long-Term Benefit Trust as parts of its broader effort to understand and manage advanced AI’s effects.

For developers and teams building AI products, the move signals that technical progress is only one part of the deployment challenge. Questions about who benefits, how systems change work, and how misuse is limited can shape product requirements, safety reviews, and public trust just as much as model capability.

Anthropic has launched a website where people can view questions already submitted and contribute their own. The initiative is also a public test of whether an AI company can turn broad social concerns into measurable commitments rather than treating them as abstract messaging.

Source: Anthropic News