Warning of ‘risk to our survival,’ pope calls for global treaty on AI

“The goal of regulation, naturally, should not only be the prevention of harmful practices but also the encouragement of best practices, by stimulating new and creative approaches and encouraging individual or group initiatives,” Francis wrote.

The statement, issued to commemorate the 57th World Day of Peace, reflects a pope who has sought to serve as a moral compass on weighty, timely issues beyond traditional religious teaching.

Although the Vatican may lack the power and influence it wielded in the past, observers insist that Francis’s focus on climate change has brought the moral imperatives of that issue into sharper spiritual focus. They suggest that the pope’s sustained focus on AI could do the same for a technology with almost limitless capacities.

Pope Francis scolds U.S., ‘irresponsible’ Western lifestyle in climate plea

“I see fear circulating among people, including lay ones, about the need for rules and safeguards,” said Vito Mancuso, a Catholic theologian and writer. “So I think, and I hope, [what the pope said today] matters. This message is not about conversion, or changing minds, but only about safeguarding humanity.”

For the Vatican, the power of the technology hit home earlier this year, when a deepfake photo of Francis strutting in a chic, snow-white puffy jacket was shared around the world and signaled the enormous — even frightening — advances in AI-created imagery.

In what could have been a reference to technologies already being deployed in countries such as China, Francis warned of the dangers of “social control” by AI. He questioned the ethics of leaving subjective judgments — who should get a mortgage or land a specific job — in the hands of unfeeling machines.

“The vast amount of data analyzed by artificial intelligences is in itself no guarantee of impartiality,” Francis wrote.

He embraced AI’s promise, too, saying the technology offered “exciting opportunities” and describing it as the “brilliant product” of humanity’s creative potential.

But he focused mostly on risks. At a time when smart drones are already being deployed on the battlefields of Ukraine, Francis called the rise of “Lethal Autonomous Weapon Systems, including the weaponization of artificial intelligence” a cause for “grave ethical concern.” A machine, he said, should not be making life-or-death decisions.

“The unique human capacity for moral judgment and ethical decision-making is more than a complex collection of algorithms, and that capacity cannot be reduced to programming a machine, which as ‘intelligent’ as it may be, remains a machine,” he said. “For this reason, it is imperative to ensure adequate, meaningful and consistent human oversight of weapon systems.”

The pope also sought to guide AI’s use, suggesting that the technology not only be applied to tackling “fake news” and “disinformation,” but also to tearing down walls between cultures to foster “fraternal coexistence.”

His words come after European Union officials last week reached a landmark agreement on sweeping legislation that could ban the riskiest uses of AI and set a global standard for its regulation. Some activists say the legislation still grants too many loopholes for the deployment of the technology by governments.

E.U. reaches deal on landmark AI bill, racing ahead of U.S.

But the European effort is leaps and bounds ahead of regulatory efforts in the United States, where bipartisan legislation has languished, and senators last week signaled that Washington would take a far more restrained approach, one focused on fostering domestic research and development. In October, a global AI summit in Britain ended with an agreement only on a vague road map for promoting safety through existing international organizations, as well as “internationally inclusive” research on the most advanced future AI models.

Francis has already met with senior executives at Microsoft and IBM to discuss the ethics of technological breakthroughs, and in his apostolic exhortation on the environment in October, he warned of artificial intelligence’s potential to become to a “technocratic paradigm” that “monstrously feeds upon itself.”