There have been times in history when the impacts of a new technology have changed the world. Perhaps that is why the Vatican, along with leaders of most of the world’s major religions, chose to host a meeting to discuss the implications for the future development of AI in Hiroshima, Japan.
Last year, representatives of the Abrahamic religions gathered at the Vatican to sign the Rome Call for AI Ethics, which IBM first signed with other industry and government leaders when it was launched by the Vatican in 2020. It is a document in which the signatories committed to pursuing an ethical approach to AI development and to promoting the human-centered and inclusive development of AI, rather than replacing humanity.
This year in Hiroshima, the Rome Appeal was signed by representatives of many of the major Eastern religions, and former signatories such as IBM reaffirmed their commitment. It was organized by the Vatican Renaissance Foundation, in collaboration with the Pontifical Academy of Life, Religions for Peace Japan, the Abu Dhabi Peace Forum of the United Arab Emirates, and the Commission for Interfaith Relations of the Chief Rabbinate of Israel.
IBM Senior Vice President and Chief Research Officer Darío Gil, who also attended last year’s event in Italy, represented IBM at the event. “AI is a horizontal technology that has implications for every industry, every country, and every value system. We believe it needs to be developed through the collaboration of many diverse institutions,” Gil said in a speech at the event. “Because more than any other technology we have invented, AI is just a reflection of its creators. And it needs to be governed like us too.”
IBM was there from the early days of AI, Gil reminded attendees. IBM co-hosted the landmark conference with Dartmouth in 1956, where researchers first tackled the topic, and it has been contributing to the field ever since.
For Gil, as for many others, the major turning point in AI came with the advent of foundation models. With AI models that can be tuned to perform different tasks with relative ease, we have seen the explosion of generative AI and, more recently, AI agents. But without ethical and open development of these tools – with as broad a community as possible to participate – large swathes of the population will be left behind by the AI revolution.
That’s why IBM launched its Trust and Transparency Principles in 2018. It was also one of the first major companies to create an AI Ethics Board to govern internal processes, tools, guidelines, training, and risk assessments for the company’s development and use of AI. This approach has been embedded in the foundations of AI development across IBM. It’s at the heart of watsonx, IBM’s data and AI platform. Its data governance and tools were designed with the same ideas and minds who work within the AI Ethics Council and can be found in the recent open source Granite family AI models that IBM has built. The Granite LLM model was recently recognized by Stamford as one of the most transparent main models available today.
This commitment to an open approach to AI development was also a driving factor in why IBM co-founded the AI Alliance As of December 2023, the Alliance now includes more than 100 companies, academic institutions, government agencies, and research labs from around the world, including Meta, Sony, NASA, Harvard, and the Cleveland Clinic. The Alliance aims to accelerate open source innovation to improve trust in AI to ensure it benefits our entire society. “We are able to work together because of our collective belief in the power of what a diverse set of institutions can accomplish,” Gil said.
Acknowledging that much more needs to be done to spur innovation in a way that leaves no one behind, Gil was optimistic about the future of such a meeting. “Much of human history has been shaped by differences and disagreements over faith,” Gil said. “But on this issue, the world’s major religions are coming together.”
