The Global Partnership on AI, an alliance of 29 members, has unanimously adopted the New Delhi declaration, pledging their commitment to a collaborative approach for AI applications that benefit people and create a global governance framework for safe and trusted AI.
The declaration also seeks to ensure that AI and its benefits are inclusive and available to all the countries of the world, including the Global South, Minister of State for IT and Electronics Rajeev Chandrasekhar said at a briefing during the ongoing Global Partnership on Artificial Intelligence (GPAI) Summit 2023.
The declaration marks a significant step forward in international collaboration on the approach to AI at a time when artificial intelligence is becoming mainstream, with intense global debate around its transformative potential and risks.
"We had a 4.5 hour ministerial council meeting where 29 partner countries of GPAI participated and have unanimously adopted the New Delhi declaration, which promises to position GPAI at the front and centre of shaping the future of AI in terms of innovation, and for collaborative AI between partner nations to create applications of AI in healthcare, agriculture and many other areas that concern all of our countries and people," the minister said.
Chandrasekhar said member nations agreed that GPAI will play a central role in shaping the future of AI governance and lead conversations on safe and trusted AI.
GPAI's 29 members include the United Kingdom, the United States, the European Union, Argentina, Australia, Belgium, Brazil, Canada, Czech Republic, Denmark, France, Germany, India, Ireland, Israel, Italy, Japan, Mexico, the Netherlands, New Zealand, Poland, the Republic of Korea, Senegal, Serbia, Singapore, Slovenia, Spain, Sweden, and Turkey.
"We have also agreed that GPAI, in keeping with the values of our partner countries, will be an inclusive partnership, which will focus on including countries from Global South and making available the benefits of AI, its platforms, and solutions to all the people of the world, including people of Global South," the minister said.
According to the declaration, the members have affirmed their commitment to continue to work within their respective jurisdictions to advance safe, secure, and trustworthy AI, including through the development of relevant regulations, policies, standards, and other initiatives.
In the last 12 months, the world of AI has transformed in a "deep tectonic manner", the minister noted.
"There is serious consensus that we have to move fast as countries in the direction of deploying AI in exploiting the technology for the common good of people, and at the same time we deal with the safety issues, the risks and the harms, and GPAI should play an important role in establishing what can be a global understanding of how to deal with those issues and make safe and trusted AI the norm," Chandrasekhar said.
Given that AI is poised for rapid expansion, there is a need for 29 countries to work together and be nimble and responsive in driving the approach to AI.
"Almost every country today is aligned to these twin objectives of harnessing the power of AI and ensuring at the same time that AI is safe and trusted," he noted.
Over the next few months, GPAI will work together to set broad principles and guardrails and possibly look to convene a mid-term sitting to discuss those principles and guardrails.
While the members of GPAI include 29 members with a shared vision, including the UK, Japan, France, and others, the intent is to expand the membership to other 'like-minded' countries, keeping with the objectives of inclusiveness. China is not part of the Global Partnership on AI.
During Wednesday's meeting, Senegal has been elevated to the Steering Committee. India has vowed that it will work towards further expansion of GPAI.
Multiple conversations are taking place across the world on AI, including the G7 Hiroshima Declaration, and Bletchley Declaration, and GPAI -- as a grouping of like-minded countries -- will contribute to the global discussions.
"It will play a leadership and central role in how the framework of AI evolves in coming months," Chandrasekhar said.
As per the fine print of the New Delhi declaration, GPAI is invested and committed to the power of AI to do good while addressing guardrails and requirements for safety and trust.
Viscount Jonathan Camrose, Minister for AI and IP, United Kingdom; Hiroshi Yoshida, Vice Minister for Policy Coordination, Ministry of Internal Affairs and Communications, Japan; and Jean Noel Barrot, Minister for Digital Transition and Telecommunications, France, were also present at the press briefing.
The declaration acknowledges the need to harness new opportunities and mitigate the risks arising from the development, deployment, and use of AI.
"This includes concerns around misinformation and disinformation, unemployment, lack of transparency and fairness, protection of intellectual property and personal data, and threats to human rights and democratic values. We further acknowledge the need for equitable access to resources, which must be considered, accounted for, or addressed in order for societies to benefit from and build competitive AI solutions," it said.
The declaration underscores GPAI’s pivotal role in addressing contemporary AI issues, including generative AI, through applied AI projects aimed at addressing societal problems and global challenges, maximising the benefits, and mitigating associated risks.