The Rome Call for AI Ethics: Academic Experts, Tech Leaders and Religious Representatives Unite in Hiroshima

Earlier this month, representatives from 11 of the world’s major religions met with academic experts and tech leaders for an AI Ethics for Peace conference in Hiroshima. Collectively, these religious leaders represent billions of faithful around the world.??

Technology and religion both have the ability to transcend borders. Religious leaders are a critically important partner to have at the table for these discussions, both because of their wide-ranging influence—an estimated 85% of the world’s population identifies with a religion—and because of their millennia-long explorations of ethics.??

Five years ago, Microsoft Vice Chair & President Brad Smith and Pope Francis met to discuss the ethics of AI. Months later, Microsoft became one of the first signatories of the Rome Call for AI Ethics, designed to promote an ethical approach to artificial intelligence and a sense of shared responsibility among governments, civil society, and the tech sector.??

At the Hiroshima conference, the Rome Call gained 16 new signatories, and Father Paolo Benanti introduced a new addendum that focuses on the governance of generative AI. In a message sent to attendees of the conference, Pope Francis said: “As we look at the complexity of the issues before us, recognizing the contribution of the cultural riches of peoples and religions in the regulation of artificial intelligence is key to the success of your commitment to the wise management of technological innovation.”?

Hiroshima served as an especially poignant setting for a reflection on AI ethics for peace. Seventy-nine years ago this August, Hiroshima and Nagasaki became the first and last places atomic weapons were used as weapons of war, causing unfathomable devastation to hundreds of thousands of Japanese civilians.??

As Microsoft President Brad Smith highlighted in his remarks at the conference, every technology can be used as both a tool and a weapon. While the atom was harnessed as a weapon before it became a tool, AI is the opposite: created as a tool, AI must be protected and regulated so that it doesn’t become a weapon.??

After atomic bombs were dropped on Hiroshima and Nagasaki in 1945, the world came together to ensure that this devastation didn’t happen again. For the past 79 years, this effort has been successful. We have the opportunity to do the same with AI.??

At the Hiroshima conference, Microsoft President Brad Smith laid out what he sees as three fundamental goals of the Rome Call, goals that also drive Microsoft’s responsible AI work.??

1) Ensuring that we bring the benefits of AI to everyone. Thomas Edison invented electricity in 1879 for a relatively narrow purpose: powering a light bulb. In the next twenty years, this invention spurred a flurry of innovation that created a wide range of technologies we rely on today, from the electric oven to the electric fan. Yet more than 150 years later, nearly 700 million people still don’t have access to electricity. We can and must do better when it comes to AI.??

2) Focusing on what can go wrong. To advance AI responsibly, we need to think not just about what might go right with AI, but also what might go wrong. Tech companies, civil society, and governments need to spend time thinking about and guarding against ways in which malicious actors might misuse AI technology. AI has already been weaponized by those who want to mislead voters and by bad actors who want to defraud people of their money or make life more difficult for women through abusive synthetic content like non-consensual intimate imagery. At Microsoft, we’re working with leaders from the world’s major religions to anticipate and solve these problems.??

3) Preserving the planet by prioritizing sustainability. Pursuing AI technology is a resource-heavy endeavor. While Microsoft is working to advance the sustainability of AI by optimizing datacenter energy and water efficiency, advancing low-carbon materials, and improving the efficiency of AI and cloud services, the climate crisis represents an urgent and existential threat that AI can also help us address. AI is already being used to forecast degradation of solar panels, track whale migration for conservation purposes, and map the melting of glacial lakes. If we want to slow the pace of human-caused climate change, we must accelerate these and other efforts.??

In June, when Pope Francis became the first pontiff to address the G7, he shared a message focused on keeping humans at the center of AI. “We need to ensure and safeguard a space for proper human control over the choices made by artificial intelligence programs: Human dignity itself depends on it," the pope said. The AI Ethics for Peace conference at Hiroshima, and the broader Rome Call efforts, are a reminder that because we all have a stake in this game, we should all have a seat at the table.??