Restraint is easy to promise and costly to practice. Its value to society depends on whether it can be seen, measured, and relied upon.
On October 21, Anthropic CEO **Dario Amodei—honored by BGF as an America 250: AI Pioneer—**defended the company’s safety record against criticism that competition had weakened its commitments. He cited commercially costly decisions, including withholding broad access to Mythos-class models for about four months while strengthening cyber safeguards. Critics, including journalist Sigal Samuel, asked for evidence that this strategy actually improves industry behavior. A company’s account of its sacrifices is a starting point for scrutiny, not independent verification. (thenextweb.com)
European Commission President **Ursula von der Leyen—BGF’s 2020 World Leader for Peace and Security honoree—**approached the issue from public leadership. In her September 16 State of the Union address, she proposed discussions with leading frontier laboratories about managing the pace of AI development. Six days later, UN Secretary-General António Guterres called for responsible pacing and credible, independent oversight. (luxembourg.representation.ec.
For the Boston Global Forum, these interventions raise a practical challenge: how can voluntary commitments withstand competitive pressure? The answer requires independent evaluation before release and continuing oversight after deployment, transparent thresholds for intervention, and obligations proportionate to demonstrated risk. Public authorities must also ensure that safety requirements protect people while preserving opportunities for innovation and new entrants.
Anthropic’s roadmap sets January 15, 2027, as the target for publishing a policy proposal on the pace of AI development—or explaining why it has decided not to. This is a milestone for public scrutiny, not a commitment to binding industry rules. (anthropic.com)
Who may set limits on frontier AI, on what evidence, and under whose oversight? These are questions a Constitution for Humanity in the Age of AI should help answer—while protecting innovation, public accountability, and human agency.
European Commission President Ursula von der Leyen delivering the State of the Union address. Photo: European Commission – European Union