GPAI: A Multistakeholder Initiative on Trustworthy AI | IGF 2023 Open Forum #111

Alan Paic

The Global Partnership on Artificial Intelligence (GPAI) is an international initiative focused on promoting the responsible adoption of artificial intelligence (AI). Established in 2020, GPAI currently has 29 member countries. Its mission is to support and guide the ethical and trustworthy development of AI technologies.

GPAI operates through a multi-layered governance structure comprising a ministerial council, executive council, steering committee, and a multi-stakeholder experts group. The ministerial council convenes once a year, while the executive council meets three times a year. The current lead chair of GPAI is Japan, with India set to assume the chairmanship in the future. This multi-level governance approach ensures that decisions are made collaboratively and with diverse perspectives in mind.

Project funding for GPAI is obtained through various mechanisms. Initial funding was provided by France and Canada, with additional contributions coming from GPAI pooled seat funding. In-kind contributions from partners and stakeholders are also welcomed to support project funding. This approach allows for a diverse range of contributions and promotes broad participation in GPAI initiatives.

GPAI is actively involved in a global challenge aimed at building trust in the age of generative AI. In collaboration with multiple global organizations, GPAI has structured the challenge into three phases: identifying ideas, building prototypes, and piloting and scaling. This global challenge seeks to address the proliferation of fake news and the growing threats to democracies. By fostering trust in generative AI, GPAI aims to ensure that AI technologies contribute positively to society.

Alan Paic, a strong advocate for GPAI, provides an in-depth overview of its governance, membership, and initiatives. His support for the project reinforces the importance of responsible AI adoption and the need for international cooperation to address the challenges associated with AI technologies. Paic also promotes the upcoming global challenge, highlighting the importance of building trust in AI systems.

In addition to GPAI, the Global Partnership on AI (GPI) has made significant contributions to the field of AI. GPI has played a pivotal role in regulating detection mechanisms in AI companies, emphasizing the importance of accountability and transparency in AI technology. The impact of GPI’s efforts is evident as some countries have incorporated GPI’s guidelines into their legislation.

Looking towards the future, GPI envisions becoming the global hub for AI research and resources. To achieve this, GPI aims to pool together global resources and expertise in AI. By bringing public research institutions together and collaborating with international networks, such as the worldwide LHC computing grid, GPI seeks to enhance understanding and advancements in AI technology.

In conclusion, both GPAI and GPI are major international collaborations that aim to promote responsible AI adoption, build trust in AI systems, and address the challenges posed by AI technologies. With their multi-layered governance structures, project funding mechanisms, and involvement in global challenges, these partnerships are crucial for shaping the future of AI in a responsible and ethical manner.

Audience

The analysis of the speakers’ statements reveals several important points regarding the Global Partnership on Artificial Intelligence (GPA) and its work. During the meeting, the GPA presented and discussed its work streams, generating significant interest. It was particularly noteworthy how these work streams were mapped against the Hiroshima commitments, underscoring the relevance and alignment of the GPA’s activities.

In addition to the mapping exercise, there was a request for insight into the GPA’s future work and thoughts on partnerships. This emphasizes the need for ongoing collaboration and clarity regarding the GPA’s direction and objectives. The speakers expressed a neutral stance on this matter, seeking more information and guidance.

The GPA’s efforts to address concerns and challenges in the field of artificial intelligence were highlighted. This includes ongoing interactions with a council that funds various projects. The council funds both ongoing and completed projects that aim to advance AI, with reports on project progress available on the GPA’s website, ensuring transparency and accountability. Additionally, the GPA seeks advice from experts in various fields to ensure the quality and relevance of their projects.

Gender diversity and equality in AI emerged as significant concerns during the meeting. Paola Galvez, a gender advocate, questioned the presence of activities related to creating diversity and gender equality in AI. This raised an important point about the need for inclusivity and addressing the gender gap in the field.

India expressed optimism about leading the GPA in the future and raised the question of whether there would be an initiative to bridge the gender gap in AI. This indicates a willingness to take action and promote gender equality within the GPA’s activities.

Peru, as the first country to have a law on AI for social purposes, expressed interest in becoming a member of the GPA. This demonstrates the broader international appeal and recognition of the partnership’s significance in advancing AI policies and governance worldwide.

Slovakia, a non-member of the GPA, is considering membership and seeks further information. Specifically, they are interested in understanding the specific regulatory support activities of the GPA and how non-members can participate in the upcoming India summit. This suggests a growing interest and potential expansion of the partnership’s membership.

The analysis also highlighted the issue of fragmented and limited public sector research on AI. The majority of research and development is concentrated in a few large private companies. This underscores the need for increased collaboration and knowledge sharing between the public and private sectors to ensure a more comprehensive and well-rounded understanding of AI.

The Global Partnership on AI (GPI) aims to address this fragmented approach by pooling resources and establishing partnerships with other international networks. The goal is to leverage the collective expertise and resources of all countries to have a greater impact on AI research and development.

Civil society and think tanks expressed keen interest in participating in the upcoming summit, showcasing their desire to contribute to the discussions and exchange of ideas. This indicates the increasing recognition of the importance of diverse perspectives and input in shaping AI policy and governance.

Finally, Ben, an advisor to the Westminster Foundation for Democracy, noted the challenges and opportunities posed by AI in election processes in the Indian presidency. This highlights the need for careful consideration and the development of strategies to address potential risks and harness the benefits of AI in these critical democratic processes.

In conclusion, the analysis of the speakers’ statements reveals various important points regarding the Global Partnership on Artificial Intelligence. The mapping of work streams against the Hiroshima commitments generated interest, while questions were raised about future work and partnerships. Gender diversity, membership expansion, public sector research, civil society involvement, and AI in election processes were also discussed. These insights emphasize the need for collaboration, inclusivity, and thoughtful governance in shaping the future of AI.

Kavita Bhatia

During the discussion, the speakers highlighted India’s vision for artificial intelligence (AI) and its potential to drive social and economic growth. They emphasized the importance of AI in bringing efficiency to administrative procedures, which in turn could contribute to economic growth. By automating various tasks and processes, AI has the potential to streamline operations, increase productivity, and foster innovation.

Furthermore, the speakers discussed how AI could empower citizens by providing them with easier access to their entitlements, thereby contributing to social growth. AI has the potential to bridge gaps and provide services to citizens more efficiently, improving their overall experience. This inclusivity was seen as crucial, particularly in a country like India that boasts a diverse linguistic landscape. The speakers stressed that AI should be inclusive and enable citizens to access services in their vernacular languages. In support of this, they highlighted the creation of a multi-modal AI platform called ‘Bhashani’, which facilitates speech-to-speech machine translation in multiple languages.

The discussion also delved into the significance of skilling initiatives in preparing for an AI-driven future. Efforts to inculcate AI knowledge at the school level were mentioned, underscoring the need to equip students with the necessary skills and knowledge to navigate the evolving technological landscape. The availability of financial support for PhD students in the field of AI further highlighted India’s commitment to fostering expertise and innovation in this domain.

The need for a Global Partnership on AI (GPI) was brought to the forefront during the discussion. The speakers emphasized the importance of GPI as a central point of contact for AI-related information, standards, and frameworks. India’s involvement in the GPI was highlighted, with the country taking the lead chair and hosting the upcoming summit in December. The aim is for GPI to have an independent identity, similar to that of the World Health Organization (WHO) in the field of health.

Finally, the speakers emphasized India’s AI approach of democratizing access to AI resources. This involves streamlined access to high-quality datasets, which are vital for research and innovation. Additionally, India aims to ensure access to compute power and skilled resources, acknowledging the significance of these factors in driving AI development.

Overall, the discussion highlighted India’s comprehensive vision and approach towards AI. By focusing on inclusive AI, skilling initiatives, global collaborations, and democratizing access to resources, India aims to harness the potential of AI to drive social and economic growth while reducing inequalities. The insights gained from the discussion underscore the need for a holistic and collaborative approach towards AI adoption and development.

Inma Martinez

The Global Partnership on AI (GPAI) has played a pivotal role in advancing the field of Artificial Intelligence (AI) between 2015 and 2018. During this period, AI experienced exponential growth and brought about significant advances in various areas, including neural networks, language models, computer vision, AI-driven drug therapy development, and level 5 automation in cars. These advancements have had a transformative impact on society.

GPAI emphasizes the importance of responsible and trustworthy AI. As AI technologies continue to evolve, there is a growing need to ensure that their development and use adhere to ethical principles and best practices in data governance. GPAI also recognizes the significance of fostering innovation in the future of work, highlighting the need to address the challenges posed by AI and promote responsible practices.

In addition, GPAI promotes the deployment of AI for industry and enterprise applications. Through a project that supports small and medium enterprises, GPAI assists these organizations in identifying suitable AI solutions for their challenges and finding local AI solution providers. This initiative aims to enhance the competitiveness of these enterprises by leveraging AI technologies.

GPAI also addresses concerns about intellectual property rights in AI. The organization has a project dedicated to this issue, recognizing the importance of creating a framework that protects and encourages innovation in AI while providing mechanisms for intellectual property rights.

The proposal to establish an expert support center in Tokyo has received positive feedback. This initiative aims to strengthen the support system for experts involved in project-based activities. Once approved, this center will provide valuable resources and expertise, further enhancing GPAI’s capabilities.

GPAI actively seeks partnerships and values decentralization to bring in as much external expertise as possible. By collaborating with research and innovation centers and specialists from various fields, GPAI ensures diverse perspectives and a multi-stakeholder approach in addressing AI-related issues.

In terms of regulatory activities, GPAI plans to organize workshops in an incubator style, covering topics such as contract laws and AIIP. These workshops, led by renowned expert Lee Tidrich from Duke University, seek to bring together specialists and encourage the exchange of knowledge. AI scientists and practitioners from any country are invited to contribute to these regulatory activities.

While acknowledging the risks associated with generative AI for democratic countries, GPAI remains driven by shared democratic values. This emphasis on democratic principles further strengthens GPAI’s commitment to addressing the challenges and ensuring responsible AI deployment.

GPAI’s projects encompass responsible AI and data governance to enhance democracy and protect human rights. The organization actively works on initiatives such as human rights projects related to data governance. By focusing on these areas, GPAI aims to utilize AI for the betterment of society, welfare, and the creation of equitable opportunities.

Overall, GPAI’s efforts in advancing AI, promoting responsible practices, supporting industry applications, addressing intellectual property concerns, establishing expert support centers, promoting partnerships, and safeguarding democratic principles demonstrate its commitment to creating a beneficial and ethically-driven AI ecosystem.

Yoichi Iida

The Global Partnership on Artificial Intelligence (GPAI) is an international collaboration aimed at promoting the responsible deployment of AI technology in society. GPAI has focused on a range of topics, including responsible AI, data governance, the future of work, and commercialization and innovation through AI technology. This comprehensive approach demonstrates GPAI’s commitment to addressing various aspects of AI and its impact on society.

To facilitate the exchange of ideas on the implementation of AI, GPAI has organized over 20 side events, providing a platform for experts, researchers, and stakeholders to come together and share their insights. These events have played a crucial role in promoting dialogue and knowledge-sharing among different actors in the AI ecosystem.

The collaboration between GPAI and other international streams has been deemed vital for achieving effective AI governance. Discussions on AI governance have been integrated into the G7 agenda, highlighting the importance of addressing the risks and challenges associated with AI on a global scale. This collaborative approach ensures that diverse perspectives and expertise are considered in shaping policies and frameworks for responsible AI development and use.

Recognising the need to strengthen GPAI, Yoichi Iida, a key advocate, believes in the significance of establishing an expert support centre in Tokyo. This centre would serve as a valuable resource by providing expert-level assistance to GPAI’s initiatives. It is noteworthy that the government is actively involved in supporting this proposal, both financially and through providing necessary personnel resources. This commitment further emphasises the importance placed on GPAI and its mission.

A proposed third expert support centre in Tokyo would operationalise the strengthening of GPAI. This new centre would play a crucial role in implementing projects and promoting the visibility and awareness of GPAI’s activities. Through this initiative, Yoichi Iida aims to enhance the understanding and perception of GPAI’s work, both within Japan and internationally.

In conclusion, GPAI is at the forefront of promoting responsible AI technology deployment in society. With a comprehensive focus on various aspects of AI and its impact, GPAI has facilitated knowledge exchange through side events and engaged in collaborative efforts with international partners. The proposed establishment of an expert support centre in Tokyo further reinforces the commitment to strengthen GPAI. Overall, Yoichi Iida’s efforts highlight the importance of responsible AI development and the need for global cooperation in shaping its governance.

Abhishek Singh

India is preparing to host the Global Partnership on AI (GPI) summit in Delhi from 12th to 14th December. The summit aims to become the leading platform for AI, bringing together nations, stakeholders, industry, and academia to discuss and collaborate on AI-related challenges and opportunities. In addition to the main themes for GPI and working groups, the summit will feature an AI expo and an AI game changers competition for startups. The deadline for startups to enter the competition has been extended from 15th October to 15th November. India is also working to expand the membership base of GPI in order to include a broader range of perspectives. They are engaging with the Secretariat and member countries to determine how GPI will be expanded. India has made significant progress in integrating AI into digital public infrastructure projects, such as the identity platform, digital payments ecosystem, and document exchange platform. They believe that AI can enhance the value and effectiveness of these projects, addressing challenges in areas like healthcare and agriculture. Collaboration is crucial for regulating AI and ensuring its fair and widespread application. India is collaborating with other nations and experts to create frameworks and guidelines for responsible AI use, addressing ethics, data governance, and other important issues. Gender bias exists in AI algorithms due to biases in input data, but efforts are being made to encourage more women to participate in AI skilling programs and balance representation. India recognizes the significance of collaboration in AI and is introducing a collaborative AI theme for the 2024 presidency, exploring shared compute infrastructure and datasets for research. While the GEP summit is primarily open to existing member countries, non-members are encouraged to participate in side events and exhibitions. India is willing to showcase their digital public infrastructure projects and AI developments to visiting countries. The country believes in sharing advancements and promoting international collaboration in the digital space. Overall, the GPI summit presents an opportunity to come together, collaborate, and shape the future of AI, with a focus on responsible and ethical development and deployment.

Inma Martinez:
Yes. Welcome, everyone, to Session 111, the Global Partnership on Artificial Intelligence, a multi-stakeholder initiative on trustworthy AI. My name is Inma Martinez. I’m going to moderate this session. I’m also participating as one of the panelists because of my role as chair of the multi-stakeholder expert group. I will make the introductions of the members of this panel. To my right, my colleague, Yoichi Iida, who is deputy director for the G7 and G20 relations at the Ministry of Internal Affairs and Communications at the government of Japan and who is also the co-chair of the Global Partnership on AI Steering Committee. And online, we have my colleague, Alang Paik, who is head of the GPAC Secretariat hosted at the OECD. And from India, we have the CEO of Digital India Corporation and India AI, Mr. Abhishek Singh, at the Ministry of Electronics and IT of the government of India. And we’re also expecting a fourth member of our panel, who is Ms. Kavita Bhatia, who is group coordinator of Emerging Technologies Division at the Ministry of Electronics and IT at the government of India. And the order of the session intends to provide you with a scope of what GPAI is as a multi-stakeholder initiative of international scope and running now in its fourth year. And each of the members of the panel in their capacity as co-chairs and as organizers of the next presidency will reflect on how GPAC is delivering value to the world, to the member countries, and what we hope for the future. And I would like to invite our colleague, Alan Paic, who is the head of the Secretariat at the OECD, who is online to start the presentation with an overview of GPAI as an organization. Alan, the floor is yours.

Alan Paic:
Thank you very much, Inma. And it is my pleasure to address you today. I will give an introduction to GPAI as a multi stakeholder initiative on trustworthy AI. So, GPAI is this long term initiative, which is specifically dedicated to AI related priorities. And it has a multi stakeholder focus to convene experts from a wide range of sectors. And the mission of GPAI is really to bring both countries on one side and experts from different stakeholder groups together to support and guide the responsible adoption of AI. And we know that especially in 2023, this becomes of very high relevance to ground, um, the adoption of AI in human rights, inclusion, diversity, gender equality, innovation, economic growth and environmental and societal benefit and seek to contribute concretely to the 2030 agenda and the UN sustainable development goals. This has been GPAI’s mission since its creation in 2020. And as I said, in this year, it became more and more evident how important this is, and how big the risks are, in, of course, in parallel to the huge opportunity, which is, which is given by the development of the technology. So we do have a global and inclusive membership, which is open to emerging and developing countries as well as developed countries. And this membership is informed by the shared principles which are reflected reflected in the OECD recommendation on AI, which is today also widespread. G 20 has also adopted a similar set of recommendation inspired by the OECD recommendations. So the GPAI members today, we number 29 members. And as I have mentioned, we do have an increasing number from emerging and developing countries, including Argentina, Brazil, India, Mexico, Senegal, Turkey, and Serbia, for example. But we also do have, as you see, most of the major leaders in in AI technology on board as GPAI members from the government side. Now, how does GPAI function, we do have a very elaborate governance, we have a ministerial council and an executive council, which are representative of the member countries, then we do have a steering committee, which is the place where the different stakeholders meet. So where we do have representatives of both the member countries and the experts. And then we have the multi stakeholder experts group of which in my is the chair. And this experts group accounts for expert working groups, and currently two expert support centers in Paris, and Montreal. And their work is also supported or will be supported in the future by the work of national institutes. So institutes which will contribute to the concrete projects which GPAI is putting forward. Now the GPAI council, as I mentioned, has two parts, it has the ministerial council, which meets once per year, and our next meeting is in New Delhi and and our colleague from India, Mr Abhishek Singh, will talk about this forthcoming summit, which is very important. The Executive Council meets at the working level three times per year and gives guidance to the GPI Secretariat on internal processes, on project financing, work plan, and approves the GPAI budget. So the Council this year, the lead chair is Japan. The incoming chair is India, who will become the lead chair for the next year. And the outgoing chair is France, who has been the lead chair last year. The Steering Committee, as I mentioned, is the place where all the stakeholders meet in this multi-stakeholder initiative. So we do have in the Steering Committee six representatives of the government and six representative of the experts. Within the representatives of governments, we have three representatives, which are the co-chairs of the initiative. We have two additional government representatives appointed by the Executive Council. And then we have a specific seat, which is reserved for an emerging or developing country, which is also appointed by the Executive Council. And this shows the commitment which GPI has to support membership from such countries. And then Steering Committee also does meet sometimes in the extended format, where all the GPI members are invited to participate in the deliberations. I think one important point is the project funding. How are the projects funded at GPI? So we do have the baseline budget envelope for projects, which has been historically provided by France and Canada, who were at the origin, the founding members of GPI. This is going to be complemented now by a mechanism through GPI pooled seat funding, where all the countries will be able to contribute. Then the second part of the project funding comes from in-kind contributions through National AI Institutes, who can contribute in-kind computing power, data resources, human resources, et cetera, into GPI projects. And then finally, we do have also partnerships, and partnerships can be from both in-kind and financial contributions to specific projects in the GPI work plan. I would like to mention also the GPI webpage, which I would encourage you to visit. You will find a lot of exciting information there. You will find our new reports. We have two new reports which are featured on our webpage, which are recent and very actual. One is about generative AI jobs and policy response, and the other one is about AI foundation models and detection mechanisms, saying that whatever new foundation model is put out there needs to provide a detection mechanism which would help us identify that that text has actually been produced by an AI. You will find information about the GPI Summit. At the GMI Summit in New Delhi, we will be launching many more new reports which are just being finalized right now. So watch this space for the new exciting results from the GPI experts. And also, you will find information about the G7 commitments to advancing generative AI policy where this collaboration does include GPI. And finally, you will find information about the global challenge to build trust in the age of generative AI. We know how fragile trust is today about the proliferation of fake news and big threats to our democracies and so on. So we do want to have this global challenge. This is a big initiative which is launched in collaboration with the OECD, with UNESCO, with AI Commons, IEEE, and VDE. And we are actually right now in a call for partners. So a call to all of you who are listening in today, if you are interested to collaborate on this very exciting initiative, please do click on this partner inquiry form just very briefly to explain what this is about. It is in three phases. In the first phase, we do want to identify promising ideas, how to have policy and technology solutions for building trust in the age of generative AI. AI. Those ideas which do show potential will get some resources to build a prototype in the second cycle, and the successful prototype then will be encouraged to pilot and scale in the third cycle. So it is a very exciting new transversal initiative with global partners. Please feel free to apply and partner with us. Thank you, Inma, and over to you.

Inma Martinez:
Thank you so much, Alan. I’d like to move now to a very important aspect of what the GPAE is, which is our mission and our vision. And I would like to invite my colleague Joichi Ida to present to us GPAE’s mission and vision and really what the Executive Council has done to address the emerging challenges that AI is presenting, with the dynamism that is required during the presidency of Japan, as well as how Japan in this year has been steering the GPAE mission in a very, very tough year and conflicting year with lots of work to do, especially from the G7 perspective of having to come up with a Hiroshima process. Thank you, Joichi. The floor is yours.

Yoichi Iida:
Okay, thank you very much, Inma, for your very kind introduction. And good morning, good afternoon, good evening, everybody. My name is Joichi Ida, actually formerly Deputy Director General for G7 and G20. At this moment, I’m working as Assistant Vice Minister at the Ministry of Internal Affairs and Communications. And my work is mainly looking at the multilateral digital policymaking, including GPAE, OECD, G7, G20, and also IGF altogether. And this is a very, very busy year for us. But at the same time, we are very happy to see different frameworks being synergized with each other, not only around AI but also other digital policymaking including the data flow or infrastructure and whatever. So with regard to the GPay operation, we took the lead chair position in the late month of last year and we are working at the lead chair for 2023. GPay has a very unique structure, not only in the organizational structure as represented by Alan in his presentation, but also in the process structure. The lead chair country hosts its summit meeting in the beginning of, on the very first day of the chair tenure, so it happened late November last year and on that very day we succeeded the lead chair position from France and we put our effort to succeed the successful GPay’s work and also even promote further its very important responsibility in promoting global AI ecosystem. Actually the GPay executive council, all member countries are working together on how to promote the responsible deployment of AI technology into the society through four different working group topics which are responsible AI and data governance and the future of work and the commercialization and the innovation through AI technology in the society. So we, the government members are closely working together with the private sector experts to promote those projects in four categories and we as a lead chair, Japan wanted to promote the uniqueness of GPA and also to improve the weakness of GPA. So in the beginning of our chair tenure, we recognized through our discussion in previous years some of the challenges for GPA would be first how to strengthen the messaging or delivering our message to rest of the world and in order to achieve that we decided to elaborate the first minister’s declaration to be delivered by the ministers at GPA summit last year. That is one instrument and secondly we wanted to add the very temporary topic for AI development as AI for resilient society. So we added AI for resilient society for priority topics for GPA activities and then we also wanted to strengthen the opportunity for experts to bring their message to rest of the world through the lot of events associated to the GPA ministers summit in the same venue in Tokyo. So we had more than 20 side events where many experts presented their own views and also do some exchanges on how we could work together to promote responsible deployment of AI technologies in our society. These are the three major focuses where our lead chair presidency worked on and I hope these three emphasis contributed to some extent to the development of GPEI work this year. And we also wanted to create a stronger synergy between GPEI work and other international work streams. That is why we picked up the AI topic as part of the G7 agenda this year and we discussed AI governance, global AI governance, as part of our working group discussion and we agreed on further work to build up the global governance policy for generative AI and in order to do that we have agreed that we should work closely with international organizations including GPEI and OECD and UNESCO to understand better how generative AI would support the human society and what kind of risks and the challenges may come up and how we could address those potential risks and the challenges through the collaboration with international organizations and experts working there. So I hope I stop there but these are the efforts by Japan as lead chair presidency and I hope these efforts will be succeeded by India to lead the chair next year and of course we, Japan, continue to make a contribution to GPEI work next year and beyond. Thank you very much.

Inma Martinez:
Thank you very much, Yoichi. As co-chair of the steering committee and colleague of Yoichi throughout this year, I have to say that the Tokyo summit was really probably one of the best events I’ve ever attended. I think as a scientist myself I felt that the sessions, not just the ones that the members of the multi-stakeholder expert group presented but also the peripheral sessions from academic and for industry that came to the summit, were truly exciting and really to the point of the times because AI these days grows exponentially, not linearly, and everybody has to react in ways that are much faster than before and with better solutions and forward thinking. So thanks very much, very grateful for all the work that Japan has done for GPEI this year and I’d like to proceed and present the other pillar of GPEI, which is the multi-stakeholder expert group, which is the group I joined in 2021 when the government of Spain, one of the members, nominated me and I entered one of the working groups which was innovation and commercialization of artificial intelligence for small to medium enterprises. And last year during our elections I then took the chair of chair. This is a very singular community because it’s the first time that the AI community of not just academics but also industry, AI scientists, lawyers, civil servants, organizations working on ethics, trade unions, have come together to work together on very very specific initiatives and this puts us in a position where we serve at the pleasure of the members but at the same time the members gives us the flexibility to proposed approaches. How would we as scientists, as AI people, would deal with some of these challenges? And this synergy is what makes the GPEI truly special and a first for many governments where it comes to AI advisory. And the makeup and the fabric of the MEC is quite varied. More than half of the expert group is composed of AI scientists, true AI scientists, and then we are complemented with other people from trade unions, for civil society, from industry that also have long careers in artificial intelligence. I would say the average age is above 40 for sure, if not more. And because the membership at present is with a huge component of European countries, well the countries, the members nominate experts, that’s why we have so many European experts in the group. But we are also expanding to bring experts not just from the membership but when we work on projects that require specific skills, we invite other members of the AI community to work with us as specialists. And this is one of the keenest efforts that we have for these years to come, which is to bring the voices of developing nations, of emerging markets, of other scientists that do incredibly valuable work that complements our own work. The gender gap is not too bad, I would say. We are making huge inroads in stabilizing this 15 percent difference because one of the points of governance that we have in the in the MEG, as we call it, is to ensure representation and not just between the genders but also geographies. Every working group has two co-chairs and we try to bring people from different continents. And what is it that we do with lots and lots of calls on video platforms because all of the experts are based in their home countries. And we are supported by the two expert support centers which are CEMIA in Montreal and INRIA in France. And this is probably one of the best elements of, you know, the governance of J-PAY that is decentralized and at the same time each member, in this case France and Canada, puts forth the centers of innovation and research to our advantage, to our support. And we organize summits. We have done summits online in 2020. We did it for the first time in person in Paris in 2021. We did it in Tokyo last year and we very much look forward to the summit in India this year. And as my colleague Alan mentioned, we engage with other organizations in common goals and projects like the Global Challenge which Alan explained earlier. Now why J-PAY engaged the AI community has a very simple answer. Between 2015 and 2018 the advancements in artificial intelligence grew exponentially. We had huge advances in neural networks, in the transformer modeling within language models. We brought computer vision, AI-driven drug therapy development, level 5 automation in cars. It was a huge movement of advancement in AI that really put a new perspective to almost 70 years of artificial intelligence. And that’s why the original founding members decided we need an association, an international initiative, that will be able to understand this massive exponential growth and transformation and governments to really get on a roadmap of dealing with it with innovative approaches. And J-PAY, because it’s an initiative, has a governance that is very singular. So we have a federated governance. Every country puts their support from their local leading institutions. And every member of the J-PAY from a country perspective brings the individual and collective leadership, exactly what we do in the expert group. And we look for multi-stakeholder equity, because we know that AI has to be an AI for all. And this is why not just the council brings new members from all geographies of the world, but the experts do the same. And one of the things that differentiates our projects with other projects around artificial intelligence in the world is that our mandate, how we have been mandated by the council, is come up with solutions, come up with real actionable solutions that go beyond policy. Yes, you can advise on policy, of course you do, but bring solutions that we can implement, that we can roll out in our global markets, and also find standards for all of us to agree upon. So the way that we understood that mandate, especially in 2023, when the emergence of generative AI really brought a new perspective and enormous challenges to society and governments, was bringing the members and the experts together. So this is something very singular. We created a town hall in May, in which the experts invited all of the member countries to attend. And we explained how we, the AI community, understood the risks, but also the opportunities for language models, foundation models, generative AI in general. And it was a town hall format. Anybody could ask anything. And we made it very free, free flow, so that for the first time it was a conversation, it was interactive. In September, we hosted the first innovation workshop for members and their delegations and our experts in CEMIA at Montreal. that you do in an innovation workshop was you challenged ideas, you address if everybody understands the same when we discuss, for example, risks. Are the risks for a member country in Europe the same as one in the Americas? Do we all prioritize the same challenges in the same way? So it was bringing the existing roadmap of challenges, risks and projects that the Council and the GPA expert group had, but we put it through the lens of are we all on the same page? Have things changed that we need to modify some of these assumptions and hypotheses? So we really behaved as artificial intelligence scientists and it was really successful because everybody felt that for the first time governments and experts were working together for two days in the same room, as you can see from the pictures, throwing ideas, challenging ideas and agreeing on approaches. And the way we operate is on a four-pillar structure. The big themes of artificial intelligence, as you all know, is to ensure that is done responsibly, that is trustworthy, that it carries the ethics of the future that we want for our people. We also concern ourselves with how the future of work will evolve with artificial intelligence coming into industry and society. And then the pillar of AI is data. So of course data governance is one of the most active working groups in the MEG and innovation and commercialization of AI. AI finally is becoming a product and a service, is coming into industry, is coming to the hands of people and we need to ensure that the service level agreements, that the human centricity by design really comes with it. As one of our experts mentioned, AI should come to us in a state that is already safe, that we don’t need to make it safe because it’s dangerous. We should really strive for an AI that comes to us in the best possible state. So how do we respond to the challenges that the member governments undergo on a monthly basis I would say? Well one of the big challenges was presented to us by the Hiroshima process in May. Together with the OECD we were called upon to support the vision that the G7 had as to how we needed to act quickly and steady and on a very solid ground with regards to generative AI, advanced AI. And immediately we looked around and we realized that we were already working on absolutely all of the points that were coming out of that mandate, out of that call. So as you can see from the list, obviously taking measures to ensure that the risks are met and addressed. We also need to mitigate vulnerabilities, you know, how it comes to market, what are the capabilities and the limitations that something coming to market in inappropriate ways would create. So the way we respond to all of these objectives is in a variety way. In some ways we produce ideas such as sandboxing for responsible AI. In others we look at what Alan mentioned, you know, detectors, real tech that actually addresses, you know, fake dissemination of misinformation etc. And as you can see from these columns, the elements of risk and concern listed by the G7, we already had projects operating in the different spheres of what needs to be done. And if you want to have a deeper view of what these typical projects are, for example one of the most exciting ones which actually has been presented in the EU Parliament and in fact is being incorporated into the amendments to the AI Act of February 2020 and also presented in the US Congress is, can we create detection mechanisms in order to ensure that when this type of AI is commercialized, it already comes with detection mechanisms that either people themselves can action and test, is this thing a fake news, but also the social media platforms. And this is real tech addressing a technological problem. It’s not a policy, it’s not a framework, it’s something very like an asset. And in the innovation and commercialization we have a project that has already entered beta which is a portal that has been launched in Singapore, in France, in Germany and in Poland as the beta testbeds in which small and medium enterprises of all sectors can consult which AI solutions are appropriate for some of their challenges, for some of their gaps. And not only that but who are the AI solutions providers at the local markets. So this is another asset that is put in the hands of industry. And one of the most exciting projects that is really addressing something incredibly hard to frame is, can we make IP out of artificial intelligence? And we started this project in 2022 and it had a fantastic year 23 because we then organized workshops in conjunction and with the support of other research institutes like Max Planck in Germany and also at Duke University in Washington DC. And it’s really addressing contract law. And contract law is very hard because the way contracts are drafted and drafted is an art and they have to have you know the proper address in the proper language and really provide guarantees. So when it comes to intellectual property the contract laws are expanding and I invite you to follow this project because in 2024 we will create an incubator. So if you work in IP law and your focus is artificial intelligence please contact us because we will be running this incubator in 24. And the next steps for other projects, for example this is the one that I lead because this is the one I created when I joined, really encompass all of the nations. Agriculture is one of the pillars of our civilization. Artificial intelligence is creating prosperity and new ways to ensure that arable land doesn’t decrease and water resources are preserved and that we really can feed 10 billion people in sustainable ways. And regulating AI as well. You know what is the landscape of AI regulation across the board? How each nation is dealing with their own AI regulation? Can we find standards? This is another exciting project. As well as the future of work. The future of work is very vital because there is much misunderstanding as to what AI brings to industry and society and there’s much fear as to perhaps being relegated to a secondary role as humanity and as workers. And this is one of their working groups that has the most activity. They have projects for 2024 in which they undergo projects with university students in South America. They will see the impact of generative AI in Espanol and they will try to get down to areas where we can learn from countries in development. As well as how the working conditions of employees and workers are changing with the rise of AI within their companies. This is just a picture. I hope that you can visit our website and get familiar with the rest of our projects. And like I said before we are a growing organization and the strength of the collective is because of what the individuals bring to us. And now it is my great pleasure to introduce to you our future presidency lead in JPEG which is the government of India. And for that we have with us our colleague in the steering committee Mr. Abhishek Singh. I believe he’s online from India. The floor is yours. Oh yes he’s there. I think I need help from the AV team in managing. I should not be the big screen now. Can you please give the floor to Mr. Abhishek who is online. I can see him and disconnect me. He’s on mute. Okay. I think Abhishek you are now able to speak. A little bit. Abhishek could you speak louder. We don’t hear much. No nothing comes through. Okay. How about now. Now you are. are. Thank you so much and for your patience. Thank you. Thank you. Thank you. I don’t know

Abhishek Singh:
what the glitch was, but anyways, but what I wanted to say is my colleague, Kavita Bhatia, is here. She will be making a presentation with regard to the summit that we’ll be hosting in December. So Kavita, can you share the slides and make the presentation? Yeah. Good afternoon,

Kavita Bhatia:
good morning, and good evening to all of you. I’ll just share my screen. Is the screen visible? Is the screen visible? Yes, yes. Yes, Kavita, it’s visible and you could switch on your camera also maybe. I don’t think I’ll be able to because this running from the one to… I’ll do it afterwards. So India’s vision for AI has been very important is that we understand that this technology brings a lot of focuses on emerging technologies, but we are wanting to ensure that the technology will bring social and economic growth, which will be in line with the inclusive development. In fact, our Honorable Prime Minister has been always saying that we need AI, we need to make AI in India and make AI work for India. And he also believes that the technology should be rooted with the principles of sabka saath sabka vikas, which means that benefit for everyone and working with everyone. So this is the basic vision of India for AI. In fact, we have a very simple approach because you all know that India is a very large country and we know that AI has the potential to improve the public service delivery by overcoming the administrative efficiency, by bringing more efficiency in the administrative procedures and keeping citizen in the focal point so that the services which we develop with the help of AI should be beneficial to the citizen. And also AI needs to overcome the traditional barrier of not having inclusivity. So we want to have it more inclusive for the development of large scale social transformation solutions. As I said that AI enables the policy developers so that they are able to take the right decision based on the data. So that the decisions taken for the development of social benefits should be meant for the citizen and he should be able to get whatever the benefit he’s entitled to. In fact, this is not what AI should stop at. It should also empower the citizen so that he knows what is his entitlement and what are the benefits which he’s supposed to get. And he should approach the government so that he’s not debarred or not having these benefits which he’s supposed to get. And the AI provides the innovative models for the governance so that we can have innovation for public’s good and create new economic opportunities. So this is the main focus and the approach which India is taking forward. And we already have come up with a strategy for AI which basically focuses on democratizing the access of AI resources. The AI resources which we mean is the streamlined access to good quality data sets for research and innovation, having the access to the compute and most important, having the skilled resources so that the innovation can be brought out into the system. So with this principles in our background, we have come up with a comprehensive program on AI which focuses on these three pillars as well as one of the most important pillar which we have kept in this program is the National Center on AI where we plan to implement 10 solutions across the country so that we can see the benefit what AI brings for the nation. Responsible AI is also one of the pillar as I said, which is very important because the solution should not cause harm to the human being. So we have already detailed out the principles of responsible AI, we have worked on the operationalization mechanism, and we have also gone ahead with one of the use case of using these principles in the facial recognition technology. As I said that India is a very large country, we have 22 languages which we speak and more, around 1000 plus dialects. So we understand that AI should bring the inclusivity and should also enable the citizens to get the services in the vernacular language. So we have already created a platform which is a multi-modal AI platform called Bhashani which is built to speech-to-speech machine translation system. In fact, in G20 we have showcased this solution and we have also added 10 international languages which has been showcased in G20. The other important aspect which we have worked on is the framework for the fairness assessment. Because the solutions which are brought to the market or which are taken for the implementation should be made sure that they are fair and should not have any biases attached to it. So we have come up with the framework and along with BIS we are also working on the other standards which are very important for developing a successful AI solution. Skilling, as I said that we have already made a note of it and we understand skilling is very important. So we are trying to cater to the skilling aspects at all the levels. One, at the very initial level when a child is in the school, we wanted to make him understand about what AI is so that we can demystify the harm which he has been told that the AI can bring. The second level is that we want to re-skill and up-skill our IT professionals so that they are up to the mark in the era of AI. So that they are able to tackle the problems which AI might bring about the job. losses. And the third area which we have tackled is the researchers. We understand that we need to have more researchers so that we can develop our own LLMs. So we have come up with a program where we are financially supporting the PhD students in the area of AI and emerging. So we are trying to cater to the skilling aspects in all the three levels so that not a single layer is being left out. With regard to the principles, with regard to the vision of AI, this year we are going to take the lead chair in December. However, we have started working for GPI as an incoming chair and we will be hosting our annual summit in December which has been talked about by Inma as well as Ellen. That December we are going to have a summit where we will make global policy makers to come and have more discussions on the responsible AI. So our main focus as an incoming chair this year has been to increase the member and the expert collaboration and in this regard we have already had convenings on three of the working groups which Inma has shown, data governance, innovation and commercialization, responsible AI, where we have brought the ecosystem of AI to understand what the GPI experts have been doing and we also wanted the industry to come and share their experiences and vision which they want from the GPI. So this we have already done and fourth convening we’ll be having shortly on future of work. The most important thing which we plan to do it in our presidency is that we wanted to make GPI as an independent identity, as a multilateral initiative so that the GPI can be the point of contact for all the AI related information or standards or frameworks as what WHO has been in. case of health. This is what we want to do in our presidency. And we also want to enhance our advocacy efforts so that we will bring more visibility of GPI outputs and also we would like the GPI work which they have done so that we can proliferate and adopt those work which the GPI has been doing in the last four years. And also one thing which we want to increase participating but we wanted to enhance this membership so that we can bring wide variety of experts which have a different national and regional views and experience so that we have a holistic view of AI in the world across. And last but not the least we want to promote the equitable access to the critical AI research and innovation which is compute, data, algorithms, software, experts and other related resources to the countries which don’t have the access to so that we have this equitable AI research and innovation access to everyone. So this is our vision for our presidency and we will be hosting a summit in December in India in New Delhi. With this I would like to thank all of you and will come out of my presentation. Thank you.

Abhishek Singh:
Thank you so much. Thank you Kavita. Thank you Kavita. Just a minute Inma. Thank you Kavita for laying down the vision and the plan that we have for the GPI summit. We are looking forward to hosting all of you in Delhi from the 12th to the 14th of December and as Kavita mentioned like apart from the focusing on the themes for the GPI and the working groups and getting all the stakeholders all the experts coming here to join it. We’re also having a few other add-ons to the summit in which like there will be AI expo in which we are getting startups from all across the world to come and show their AI based solutions that they are there. We’ll be having an AI game changers that we have launched. We have shared the information with all of you. are startups who are building any solutions related to AI, any dimension of AI. If they want to participate in that, the last date of it has been extended from 15th of October to 15th of November. We would like you to share it with the relevant stakeholder community with regard to the AI game changes. And we’d also have a lot of side events which will focus on various themes of AI. So if there are any member countries or any of the stakeholders who are represented at the IGF would want to contribute to the discussions in the side events at the GPA summit, we look forward to hearing from you, look forward to getting your involvement, your participation. Because the way the summit is being planned in India is that we want to make it like the go-to, just like the Internet Governance Forum is the go-to place for issues related to internet governance that we all do. And we all look forward to such events which are held annually. Similarly, the GPA as the prime body with regard to artificial intelligence will bring together all nations, all stakeholders, civil society, non-government organizations, industry, academia to this partnership. The GPA needs to evolve into that. We are working towards that. And we are also working with the Secretariat and member countries with regard to how the GPA will be expanded. So look forward to getting all your views and getting your participation in the GPA summit in December.

Inma Martinez:
Thank you so much, Abhishek and Kavita. Thank you so much for showing us what is to come in 24. I’d like to move to the discussion now. And I want to remind our audience that we will have 15 minutes of Q&A that you can use the chat or you can make the questions here in the room in person. But first, I’d like to start with Joichi Iida. You always talk about strengthening GPA. Can you give us some highlights as to why you firmly believe in that and what it is to come in that respect?

Yoichi Iida:
OK, thank you very much for the question. And yes, we strongly believe that the strength and uniqueness of GPA exist in the expert level. structure and multi-stakeholderism is the center of the value. So ideally the government and other stakeholders should support the private sector experts who are working in the project-based activities through working group works and their work is now supported by two expert support centers located in Montreal and in Paris. So in order to strengthen the GPA’s value and function we would like to strengthen the support system to expert activities and that is why we are proposing to add expert support center in Tokyo. So we have proposed this proposal to establish and add new expert support center in Tokyo at executive council this year and we believe the proposal was welcomed generally and of course we need to go through the discussion at steering committee and ministerial council but once it is approved we would like to operationalize the concept of third expert support center. In order to do that we at the same time need to prepare on our side to bring some financial and personal resources to manage the center and we are now internally discussing across the government how we could do that. Of course we need as government to bring this in action. And that is one of the ideas we want to realize, and also we are trying to promote the visibility and awareness among people on GPAE’s activity, and we hope this will be promoted through the closer collaboration between GPAE and the Hiroshima AI process, where we are discussing how we could promote and materialize project-based activities to accumulate some evidences on what kind of measures and practices might work to address some of the potential risks and challenges brought by generative AI and the foundation models. And also we may do some project to understand better how we could responsibly deploy generative AI and the foundation model into the society. And these topics and projects can be implemented through the newly established expert support center in Tokyo. That is one of the ideas we are now trying to promote, and I hope this will contribute to the higher scaling up of GPAE function. Thank you very much.

Inma Martinez:
Thank you. Thank you so much, Joichi. I’m really delighted to hear this because, as I mentioned earlier, the strength of GPAE resides in the multi-stakeholder equity and the decentralized and federated governance. And nothing would delight me more than having an expert group in Tokyo. I’d like to ask something to Abhishek Singh. that is also super exciting because we are all looking forward to having India in the president’s seat, in the lead seat. And I would like to ask him, every time that we have met and you have come over to the working groups and you have looked at the projects, what do you feel is probably the significant difference between our projects and your agenda as a CEO of the digital agency in India, under a very visionary mandate from your president?

Abhishek Singh:
Thank you, thank you Inma. I must straightaway mention that one key value that we get as being part of the GPA and getting to interact with the multi-stakeholder group, the Center for Excellence at Montreal and at Paris and interacting with experts who are all working on various projects, the four working groups that we have, is that we get a lot of insight with regard to what more can be done. As you rightly mentioned, India has been the leading country when it comes to implementation of digital public infrastructure projects. We’ve implemented certain digital India, as we call it, the digital India as a brand, but the digital project will be implemented at population scale, whether it’s the identity platform which has got more than a billion people registered, we have almost 70 billion. Authentication that happen on a daily basis or our digital payments ecosystem which is like one of the world’s most robust and most largest digital payments platform with more than 10 billion transactions that happen every month. So the numbers are huge or our document exchange platform which we call DigiLocker, which has more than 200 billion registered users. So in India, whatever we do is at a scale. But now as we move on and try to leverage artificial intelligence, we are seeing that a lot of value addition that happens when you kind of bring in a layer of AI to the digital transformation project that we have implemented. And when we do that, when we say, for example, use face recognition for authenticating people. We are using a very simple AI tool. But then all the issues related to ethics and responsible AI come in. When it leads to more and more adoption of AI, the future of work comes in as a very important thing. We have a large number of population, almost 50 million Indians are working in the IT and ITES sector. But the way AI is coming in, some of these jobs will be impacted. So we need to work with the global community. We need to work with the experts. We need to work with other nationalities with regard to especially coming up with framework, the guidelines for regulating AI, ensuring that innovation and regulation go hand in hand, ensuring that we are able to provide equitable access to AI, ensuring that we are able to democratize AI, and most importantly, bring in an era of explainable AI. So very often, these things cannot be done by only one nation alone or a few corporations alone. There is a significant concentration of the AI technology in a few companies and a few nations. But if we need to harness the benefits of this technology, we need to take it further. We need to ensure that there is access to compute, there are frameworks with regard to data governance, the frameworks with regard to leveraging the data that can be used for building AI models. And there are solutions that can be used for population scale societal problems. Like for example, how do we build AI solutions for solving healthcare issues? How do we use AI to detect tuberculosis or cancer? How do we use AI for helping farmers across the country? And when we do that, the real benefits of AI will come in. And the low and medium income countries are going to benefit a lot in doing that. So what we are doing is that integrating the artificial intelligence and advancement of field of AI with the digital public infrastructure project that we have implemented and work with the global community in order to fast forward that. And whatever we have learned, whatever we have built, or whatever we are building together, it becomes part of the global DPI repository. As the G20 declaration mentioned about building a global DPI repository, the AI solutions will also become part of the global. repository and many of these solutions that will be developed in world in cooperation with other member countries will become available for adoption and replication across the world. So that’s the value we see as being part of the GPA and the benefits that we get in engaging with the real experts, real technologists, real engineers and real social scientists who are working in this field.

Inma Martinez:
Thank you. Thank you so much. I completely second everything that you said and something very important that both of you have mentioned and especially in the themes of each presidency, resilient society, making empowering people to respond to challenges and then making AI equitable and accessible to all. This is the century of human centricity, putting people at the center of everything that we create so that we can really create a future for everyone. I’d like to open the floor to anyone online or present in the room to have the chance and the great opportunity to make any questions to Joichi and Abhishek or myself. So if we have any questions, please raise your hand and somebody will bring you a mic. Or let me just check anything happening online. Yes, we have one question over there. Can I? I’ll give you the mic myself. Please let us know who you are and your organization. There is switch. Great. How’s that? Perfect. Okay.

Audience:
So first I just want to say thank you so much for the great presentation. I found it really helpful. My name is Ed Teller. I’m from Amazon Web Services from the global AI policy team. I thought the slides were all great and the one which I thought was particularly interesting actually was seeing how GPA’s work streams were being mapped against the Hiroshima commitment. So I thought actually I just wanted to ask a couple of questions about that sort of firstly sort of how you see that work sort of going forward. Because you’re likely going to see reporting against those, the Hiroshima principles. So I think that’s a really helpful lens to understand the work. And also just in terms of like how you’re thinking about partnerships as well and collaboration against those work streams too. So yeah, those might be questions, but thanks very much for the really helpful presentation. Thank you.

Inma Martinez:
We have an ongoing interaction with the council. At the inception, the member countries exposed their concerns, the challenges that they felt that they needed to address. And then we looked at those concerns and challenges and from our perspective of engaged scientists, because we all work back home at our universities, labs, industry. And we decided, okay, this is how we would address that issue. And we would propose… projects and things to develop and then the council funds it and we set ourselves to deliver the specific initiative, the specific projects. Some of them are ongoing, some others were completed within a period of two years and all of this is shared back to the member countries and also publicly because all of these reports as to how we are progressing are hang on our website and our approach to the projects is we cannot be the only ones working on these. So as Yoichi rightly said, the strength of JPEI is that once we get our mandate, we look at the world and we ask ourselves, is there an expert somewhere that we should invite to work with us in this project and those are the specialists. So for example in my agriculture project, immediately I reached out to NARO in Japan who is the agency that looks at all the technification of agriculture and the director-general himself is my specialist and he comes to the meetings every two weeks and he has brought information, insights, thoughts, strategies and that’s how we work and this is how the community really brings real insights because it comes directly from the places where AI is being created and we are expanding and now we’re moving into IP law and we move to civil society and ethical societies and companies that think about service-led innovation when it comes to putting products in the world. What are the principles that you are guided by when you create a product to be safe and that’s really the uniqueness of this initiative which truly is unique because there’s nothing like it and we hope to strengthen it with the support of our member countries and their leadership and their own research institutes and innovation centers. That’s why it’s decentralized and federated. Thanks so much for the question. Thank you. We have another one here. Let me just pass you the mic.

Audience:
Good afternoon. Good morning. My name is Paola Galvez. I’m Peruvian, right now based in Paris. I just finished my graduate internship at the OECD and my master’s in Oxford. Former advisor to the Secretary of Data Transformation in Peru where I helped over, actually I oversaw the design of the AI national strategy and I have one question because personally I’m a gender advocate in whatever I do. So I saw in the chapter or initiative that you have in Responsible AI there is an activity number five creating diversity and gender equality in AI. If you please could explain or expand a bit on it. Also ask India the future. chairs, because I saw their principles and it was really fantastic to hear their optimism and how they want to position GPA as the institution or multilateral initiative to really come to when we need expertise. And if you will be open to develop an initiative that works in bridging the gender gap in AI that we have in the world. And my second question would be how, coming from Peru, I am happy to see many developing countries, but maybe what are the requirements to have other countries as members? Speaking of Peru, that we are the first one to have a law on AI for social purposes. I think that would be something that our prime minister would be interested to, but I would like to know what are the requirements so that I can come to them and tell them how incredible initiatives you guys have. Thank you.

Inma Martinez:
So the gender question is for me or for Abhishek?

Abhishek Singh:
I can take that, no worries. Thank you, Abhishek. The floor is yours. You can give your question. Yeah, thank you. In fact, I really thank, I didn’t get your name, the lady from Peru for the question, but a very rightful and very impactful issue because gender has been a issue, especially with regard to AI algorithms and the biases in AI, whichever way we work in. And that’s primarily because of the bias that exists in the input data. And if the data is not equitable, in the sense that if data is biased, like for example, very typical examples are given that AI models would treat engineers to be men and teachers to be women. So these biases, if we are aware of, that can be resolved to some extent. This is a very basic level. The other biases that exist when it comes to gender and AI. like very often in India, what we’ve seen is that even in AI scaling, the number of people who undergo AI scaling and AI training, very often because of societal biases, it’s the men and the boys who take the dominant share of because of access to devices, because they have access to higher education. So all those biases come in. So we have a conscious plan within India that whenever we do talk of digital training and digital scaling, we try to balance it out and we try to have proactive measures in order to encourage more women to take up courses and take up AI-based scaling, so that ensures that we have a fair balance of that. Whichever data sets are used, how do we mitigate the biases which are there in order to ensure that AI is gender neutral and AI is more equitable for the whole population? So those are some measures which are taken, but yes, it will take a lot of time to train the models in order to ensure that they’re aware of the biases and how to get rid of them. And that’s the work that’s part of the ethical framework and the responsible assessments of AI solutions, wherein we address the biases coming in for gender or biases coming for race, or biases coming for other diversity that exists in the world. So that’s on the gender issue. And the collaborative part is, again, something which is very, very useful. And one of the themes which we have introduced for the 2024 presidency is with regard to collaborative AI for building partnerships amongst various stakeholders. Like, how do we join hands? How do we share knowledge, share experience, share models, and together work on building big solutions? In fact, Alan from the Secretariat very often talks about building a CERN for AI. Just like world community has come together to work for particle physics and advancement in the CERN Center, can we think of having a shared compute infrastructure? Can we think of having shared datasets on which research could be done? Can we think of sharing insights or sharing partnerships between AI research? from across institutions. So that would really, really ensure that we work together collaboratively in the field of AI and develop AI for the betterment of humanity, rather than just being always being aware of the biases or the bad things that can come from artificial intelligence.

Inma Martinez:
Thank you. I feel the second question about further countries joining. Right, maybe. Maybe Alan. Okay. Hello.

Alan Paic:
Yes, it was not about further countries joining. Well, I can also mention that. So we do have a membership process, which is well-defined and also described on the website. Further countries are invited to apply. Right now, the intake for 23 has been closed. And there will be a next opening in 24. So watch the GPI website. Everything is explained there. The deadline is around June or July, and countries are expected to present a letter of intent with their motivation to join GPI as an initiative committed to trustworthy AI. So that is the possibility for membership. And I would also like to react to what Mr. Singh just mentioned. Yes, we have been talking about the future perspectives of GPI. You know, GPI has achieved a very significant impact and Inma has mentioned previously, for instance, the detection mechanisms, the obligation by, you know, companies putting on the market foundation models to actually provide the detection mechanisms, which would allow us to understand that something has been produced by that foundation model. That is something which is very significant and has already been taken on board in some countries’ legislations. And I think GPI going forward wants to provide more and more impact. And I’m really happy to hear that India has this vision for the new year and new presidency to lead the way of GPI towards, you know, pulling together the resources. Basically, the idea comes from the understanding that today a lot of the R&D is concentrated in the hands of a few. large private companies, and that the public sector research is far behind and has a very limited understanding of the new technological advances, and the spending from the public sector is fragmented among different countries. So the idea, as Mr. Singh mentioned, of GPI as a go-to place, as a place where all the countries come together and pool together their resources in data and in compute power, and also with perhaps some other international networks which already exist, such as the worldwide LHC computing grid for particle physics, this could bring GPI even further on this ambition. So we do, of course, want to partner with private companies, that is great, but we do also want to bring together public research institutions, and this is the model of the national AI institutes which I mentioned in my introductory speech. Thank you.

Audience:
Thank you so much. I think we have a question here. Thank you very much. My name is Juraj Corba, I’m representing the government of Slovakia from the EU. I have two questions, if I may. In your presentation, which was very helpful and very informative, thank you for that, you have mentioned that you are planning some kind of activities for the support of regulatory activities. Could you please be more specific? What are you planning? So that’s the first question. And the second question relates to the India summit, so it may be a question to our friends from India. My question would be, you have mentioned that you are planning to create or manage the summit in a way that it is the place where we go and come for the AI topic. Slovakia is a non-member of GPI, we are considering a possible membership, but my question is how open will you be to participation of non-member countries, and how they can effectively participate at the India summit? Thank you very much.

Inma Martinez:
So I will answer the first question, and I believe the second is for Mr. Singh. So these workshops, what I mentioned regarding the contract laws, regarding AIIP, are planned to take place in incubator style, and if you go online and you look for this project, you’ll see the person that is leading this, which is Lee Tidrich, which is a professor at Duke University, and she will then be able to share the schedules, because as I mentioned, the projects seek specialists, and the specialists are invited to work with us from any country. So if there are projects in which you feel that some of your AI scientists and practitioners would really like to participate, we would like to invite them. to contribute, the projects are open for collaboration. And then I believe the second question was for India, for Mr. Singh, for Abhishek. Can government delegations attend the summit in India?

Abhishek Singh:
Yeah, thank you, thank you. I got the question, sir. So thank you for the questions, Slovakia, and for the interest in the summit. So while the way the GEP is constituted, the ministerial and the various official engagements which will happen on the 13th and the 14th will only be open to the existing member countries. And of course, the membership application, I don’t know how quickly we can work about approving the membership, but yes, there are the side events on the 12th, and there will be an exhibition which will be open to all guests, all members. So if you’re wanting to come and visit, you can write to us and we’ll work out the modalities and the sessions in which you’ll be able to participate as a non-member. So that we are willing to look into it, and we would also be showcasing, there’s a lot of interest in our prime minister himself said that if there are other countries who are coming, then we can showcase them our digital public infrastructure projects, how other parts work, how we have done in AI, and the side events and all in which we could participate. But yes, the ministerial and the steering committee and the executive committee, those formal events of GPA, they will be open only for the member countries.

Inma Martinez:
Okay, any other questions? Yeah, we have.

Audience:
It’s just a continuation to his second question. And I just introduced myself. I’m Kamesh Shekhar, I’m from India. And yeah, it’s a proud moment for us that we will be having the summit very soon. I’m from the think tank called the dialogue. We are based out of Delhi. And I just had a very follow up question to his is that like, and so it had actually mentioned. that there will be side events and other opportunities. I just wanted to understand like as a think tank and as a civil society, how can we also take part in the summit and like what are the way forward that we could watch out for as the summit like comes into picture. So that’s just a question I had, yeah.

Abhishek Singh:
So the details of side events will be up on the website very soon, hopefully by next week or so. And we would welcome registrations for the side events from non-members also. So, and there are a lot of think tanks who are already wanting to take part. There’s a lot of interest from the industry. There’s a lot of interest from the startup community. Within India, we had a big meeting yesterday in which more than 50 people had participated and they all have given various ideas with regard to what all we should be covering, especially with regard to building consensus with regard to key issues that the world faces, especially for the advancement of artificial intelligence and other technologies. So we look forward for that. We have been getting very good response from all stakeholders, especially the G20 countries and the countries beyond for our initiatives. In fact, I would like to just mention that even in the G20, as part of the Digital Economy Working Group, we, based on the requests that we got from multiple countries, we hosted a global DPI summit in Pune in June. And that included a lot of countries outside the G20. So almost 50 countries took part who were not the members of G20 because they wanted to know about what all we have been doing in the digital space. And eight of those countries already signed MOUs with us for replication of some of the India-stacked solutions in their countries. And none of them were the G20 members. So similar approach will be here. So the official meetings would be open only for the members, but the non-official, the exhibition, the side events, the keynote talks, we are trying to get some keynote AI scientists and researchers who can come and deliver a keynote talk. So those will be available for people who are not officially a member of GP also.

Inma Martinez:
Okay, and we have a question right next to you.

Audience:
Thanks very much. My name is Ben-Gurion Jones, I advise the Western Institute Foundation for Democracy and the body on issues around technology and elections. So I’m working on 10, 15 elections. I think the mic is not on. The mic is not working, is it? Can’t hear. There we go. I’ll start again. My name is Ben Graham-Jones. I advise the Westminster Foundation for Democracy, a UK public body around democracy and elections, but especially on issues pertaining to technology. I understand that you’re a big principle of GPA as well as that it’s very much guided by the shared democratic principles of its members and I’d be keen to know you’re moving forward into the Indian presidency, whether there’s also plans to address new issues around AI in the election processes, both the challenges and the opportunities moving into the year ahead.

Inma Martinez:
Well, from the MEXT perspective, we know that one of the major bruising points of generative AI and AI that has been misused is the risks to democratic countries, the risks to the democracy in the world. So these cascades into various projects, not just the one, because we believe that the pillars of the world is democracy, a welfare society that looks after its people and ensures their well-being. So if you want me to look through all the other projects after the meeting and tell you which ones, the theme is running across various projects from responsible AI to data governance. For example, data governance in 2022 had a specific project on human rights and that obviously comes from, you know, non-democratic situations. So I can talk to you after the session. Are there any other questions? I think we have reached minus one minute and I would like to thank Joichi Iida and Mr. Abhishek Singh and Kavita Bhatia and our colleague in Paris at the OECD, Alan Paik, for convening and being with us and presenting our vision, our hopes for the future and the singularity of this initiative, which many times when people ask me what the GPA is, I always say when you wonder if governments care about people, this is one of those. They truly do, they truly do, and they do their best to really take the reins of our future and make sure that AI is for opportunity and for good and for welfare. And let’s hope that we can achieve that. Thank you all for coming to this session. Many thanks to those who connected online, and I declare the session finished. Thanks very much for your questions. Thank you, thank you very much. Thank you.