Anthropic co-founder Chris Olah said Monday that artificial intelligence development cannot be left solely to technology companies [1].
This call for external oversight suggests a growing admission from industry leaders that the commercial pressures of frontier labs may conflict with global ethical standards. By advocating for a multi-sector approach, Olah is signaling that technical safety is insufficient without social and moral governance.
Olah spoke May 25, 2026 [1], in Vatican City during the launch of an encyclical on AI by Pope Leo XIV [2]. The 33-year-old executive said that AI cannot be steered by frontier labs alone [3].
He said that oversight is required from religious leaders, governments, and civil society [4]. According to Olah, AI labs face specific pressures that may conflict with the greater good—risks that include massive job losses and the potential to fuel conflict [5].
"AI development cannot be left solely to technology companies," Olah said [6].
His remarks coincided with the Pope's theological document, which addresses the risks of AI fueling warfare and calls for robust regulation [7]. Olah said that external actors are necessary to safeguard society from the rapid deployment of these technologies [5].
"We need oversight from religious leaders, governments, and civil society," Olah said [4].
Throughout the event, the discussion focused on the necessity of a partnership between tech ethics and traditional moral leadership to prevent the destabilization of labor markets and international peace [8].
“"AI development cannot be left solely to technology companies."”
The alignment between a leading AI lab co-founder and the Vatican indicates a shift toward 'human-centric' AI governance. By acknowledging that corporate incentives can diverge from the public interest, Olah is validating the need for regulatory frameworks that include non-technical stakeholders to mitigate systemic risks like mass unemployment and automated warfare.
