Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Amb Thomas Schneider
20 Feb 2026 12:00h - 13:00h
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Amb Thomas Schneider
Summary
Thomas Schneider opened the session by thanking the Indian government and the global audience for gathering at the AI Impact Summit in Delhi, emphasizing that the event’s focus on “people, progress, planet” reflects a shared ambition for inclusive AI development [1-3]. He reiterated that AI must be harnessed to deliver economic and societal benefits for everyone while safeguarding human dignity and the environment [4-6]. Switzerland announced it will host the next AI Summit in Geneva in 2027, noting the strong enthusiasm of Swiss and international stakeholders who are already contributing ideas for the agenda [8-10].
Schneider stressed that Switzerland’s motivation is not to stage a show but to make a substantive contribution to ensuring AI’s transformative power-comparable to the printing press or the combustion engine-raises global quality of life rather than diminishes it [12-14]. He outlined a plan to build on existing governance mechanisms such as the UN Internet Governance Forum, AI for Good Summit, ITU-UNESCO forums, OECD, and prior AI summits, thereby avoiding duplication and leveraging proven platforms [20-21]. Drawing an analogy with the two-century evolution of engine regulation, he argued that a complex mix of technical, legal, and societal norms-already evident in transport and manufacturing-must be developed for AI [27-34].
The speaker highlighted the Vilnius Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law as a principle-based framework that can be adapted globally, offering flexibility for diverse legal traditions while promoting interoperable standards [46-49]. Concluding, Schneider called for broad collaboration, positioning Switzerland as a facilitator that will bridge stakeholders from all regions to create pragmatic, trustworthy structures that enable AI to support peace, prosperity and human dignity, and expressed anticipation for the Geneva summit in 2027 [55-58].
Keypoints
Major discussion points
– Inclusive, human-centred AI for the benefit of all peoples and the planet – The speaker stresses that AI must be developed “so that everyone in the world can benefit” while respecting human dignity, autonomy and the environment [4-6][14-15].
– Switzerland’s facilitating role and the 2027 Geneva AI Summit – Switzerland will host the next summit, intends to “build on” existing platforms (UN-IGF, AI for Good, OECD, etc.) rather than reinvent them, and will act as a neutral facilitator that brings together diverse stakeholders [8-10][20-23][55-56].
– A multi-layered governance approach modeled on the regulation of engines – The talk draws an analogy to the historic governance of combustion engines, arguing that AI will require a “set of thousands of technical, legal, and also non-written societal norms” and a mix of binding and non-binding instruments [27-34][41-45].
– The Vilnius Convention as a cornerstone and the need for additional norms – The newly-negotiated Vilnius Convention on AI, Human Rights, Democracy and the Rule of Law is highlighted as a “principle-based framework” that can be adapted globally, while acknowledging that many more sector-specific norms will be required [46-52].
– Call for global, especially under-resourced, participation and gap-filling before the summit – The speaker pledges to help “less resourced communities” navigate the complex governance ecosystem and to use the time until Geneva to identify and close gaps in global and regional AI governance [23][42-53][55-56].
Overall purpose / goal
The address aims to set the agenda for the forthcoming AI Impact Summit in Geneva (2027), positioning Switzerland as a collaborative host that will coordinate existing international forums, advance a human-rights-based governance framework (exemplified by the Vilnius Convention), and mobilize worldwide stakeholders-including those from the global north, south, east and west-to co-create pragmatic, interoperable norms that ensure AI’s benefits are shared equitably and its risks are mitigated.
Tone of the discussion
The tone is consistently respectful, optimistic, and constructive. The speaker repeatedly expresses gratitude, enthusiasm, and a willingness to listen (“we are very keen to hear your ideas” [15]), while emphasizing partnership and collective problem-solving. There is no noticeable shift toward negativity or confrontation; the discourse remains collaborative from start to finish.
Speakers
– Thomas Schneider
– Areas of expertise: Global technology governance, artificial intelligence policy, internet governance, human rights and democracy in AI.
– Roles and titles:
– Ambassador and Director of International Relations at Ofcom Switzerland[S3]
– Vice-Chair of the Council of Europe’s Committee on Artificial Intelligence[S3]
– Former Chair of ICANN’s Governmental Advisory Committee (2014-2017)[S1]
Additional speakers:
– (None identified in the transcript)
Thomas Schneider opened the plenary by thanking the Indian government and the global audience for convening the AI Impact Summit in Delhi, and he framed the event’s three-fold focus on “people, progress, planet” as a shared ambition for inclusive AI development [1-3]. He emphasized that the promise of artificial intelligence must be harnessed so that everyone, regardless of geography, can share in economic and societal progress while safeguarding human dignity, personal autonomy and the health of the planet [4-6].
He then announced that Switzerland will host the next AI Impact Summit in Geneva in 2027. Schneider highlighted the strong enthusiasm already evident among Swiss stakeholders and the positive reactions from international partners, noting that many governments and civil-society actors have begun submitting ideas for the agenda [8-11]. The Swiss motivation for organising the next summit is “not to make a show”, but to make a substantive contribution to the global good-use of AI [12-14].
Drawing on history, he compared AI’s transformative capacity to inventions such as the printing press, radio, television, the internet and the combustion engine, arguing that-like those technologies-AI should raise rather than lower the quality of life for all peoples [13-14]. He also used an industry-regulation analogy, pointing out that, unlike the highly harmonised global airline sector, car regulations remain fragmented, underscoring that AI governance will likely exhibit similar variations across domains [60].
Schneider said the Geneva summit will carry a “Swiss flavour”, meaning it will be grounded in Switzerland’s tradition of constructive, neutral facilitation and will build on, rather than duplicate, existing multistakeholder mechanisms [18-23]. He listed the platforms that will be leveraged, including the UN Internet Governance Forum, the AI for Good Summit, the ITU-UNESCO Global Forum on Ethics of AI, OECD initiatives, the Global Partnership on AI (GPI) and other regional bodies [20-22][61].
To ensure that less-resourced communities can participate, he pledged continued support through long-standing partners such as the Diplo Foundation and the Geneva Internet Platform, enabling these groups to navigate the complex governance ecosystem and have their voices heard [23][55-56].
Schneider then described a layered governance architecture, noting that societies have never relied on a single institution to regulate transformative technologies. He illustrated this with the evolution of engine regulation, where “thousands of technical, legal and non-written societal norms” now coexist to manage transport, manufacturing and energy [27-34][41-45]. Accordingly, Switzerland is analysing existing AI governance instruments, identifying gaps, and drafting both technical norms and binding or non-binding legal instruments [42-46].
Central to these efforts is the Vilnius Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law, which he chaired negotiations for among 55 countries. The convention provides a principle-based, flexible framework that can be adopted globally while allowing states to embed the principles within their own legal traditions, thereby promoting interoperable rather than identical regulations [46-51][49-50].
Schneider acknowledged that the Vilnius Convention alone will not be sufficient; additional sector-specific binding and non-binding instruments will be needed to ensure coherence across the AI governance landscape [52-53]. He called for the period leading up to the 2027 summit to be used for systematic gap-identification, the development of pragmatic norms, and the coordination of existing initiatives so that AI can continue to drive innovation while legitimate concerns are addressed [53-54].
Emphasising Switzerland’s role as a neutral facilitator, he said the Swiss team will act as bridge-builders, identifying areas of shared vision among stakeholders from the global north, south, east and west, and translating those consensuses into workable steps [55-56]. He reiterated the importance of inclusive participation, noting that the summit will strive to create “trustworthy cooperation” that respects dignity, promotes peace, prosperity and security, and ultimately allows AI to serve humanity and the planet [57-58].
In closing, Schneider expressed confidence that the collaborative approach-grounded in historical lessons, anchored by the Vilnius Convention, and reinforced by existing multistakeholder platforms-will enable the Geneva AI Impact Summit to produce concrete, interoperable outcomes. He thanked the audience for their support and looked forward to meeting the global community again in Geneva in 2027 [58-59].
So, dear friends and colleagues from India and from all around the world, it is an honor and pleasure to be here with you in Delhi at this pivotal moment for global AI governance. And first, of course, I want to express my gratitude to the government of India for bringing together a diverse and distinguished group of leaders, innovators, researchers, civil society representatives from all around the world. Switzerland very much welcomes and supports the focus of the AI Impact Summit, which is well presented in the three sutras, people, progress, planet, as we all have learned in the past weeks and months. And we fully agree that we need to develop and use AI in a way that everyone in the world can benefit from the potential that AI offers.
This includes economic and societal social progress for everyone. At the same time, of course, we need to make sure that we are able to develop and use AI in a way that everyone in the world can benefit from the that we respect human dignity and autonomy, as well as our planet, which is the basis for all life that we know, at least so far. We haven’t found other life elsewhere. So we are honored and very proud to be hosting the next AI Summit in Geneva in 2027. It is overwhelming to see already now and feel the momentum and the enthusiasm that we sense on national level among all Swiss stakeholders, as well as the very positive reactions from our partners from all around the world, who are all eager and willing to cooperate with us and contribute to the summit in Geneva.
Already now, we are approached by many governmental and other stakeholders that share their ideas with us about what the Geneva Summit and the road leading up to it should focus on and what it should achieve. And let me assure you that this is very welcome and helpful to us. The Swiss motivation for organizing the next summit is to, not to make a show, it is to substantially and meaningfully contribute to achieving the goal that mankind and the world want to achieve. it is to substantially and meaningfully contribute to achieving the goal that mankind uses the unprecedented potential of AI to achieve the goal that mankind uses for good and not for bad. This potential of AI, which may be at least as transformative as the invention of the printing press, radio, television and the internet, as well as the invention of the combustion and other engines together, this potential must be used to raise and not lower the quality of life of all people in the world and not just a few.
AI must strengthen and not weaken the dignity and autonomy of all people in the global north, south, east and west or whatever we call the region where we live and help us all to live together in peace and prosperity. So we are very keen to hear your ideas about what we could and should do together to achieve this goal. Of course, we do have some ideas on our own, but we have not decided yet about the focus of the Geneva Summit. We will discuss it with you together, shape it together. Of course, there will be a Swiss flavor to the Geneva Summit, which is based on the way we work and what we understand, our role in the international community.
We will try to be constructive. Thank you. creative and innovative and try to find pragmatic and fair solutions through bringing together all stakeholders in their respective roles and with their respective experience and at the same time we will try not to reinvent the wheel and duplicate processes and instruments that already exist and that work but rather we will try to build on them because we do already have a number of dialogue platforms for AI governance and for sharing good practices such as the UN Internet Governance Forum and its national and regional initiatives, the AI for Good Summit and the Global Forum on Ethics of AI organized by ITU, UNESCO and many other UN related processes and forum.
We have other forum like the OECD, GPI and other international and regional organizations and of course we will build on the outcomes of the previous summit in the UK, Korea, Japan, sorry Paris, Japan will follow at some point in time, UK, Korea, Paris and of course here in Delhi and we should not forget There are many academic and other networks that provide expertise and solutions. So we will do our best to bring them all together. And with the help of our longstanding partners from the Diplo Foundation and the Geneva Internet Platform, we will also try to facilitate the orientation in this complex governance ecosystem, in particular for less resourced communities, so that also they know better about what is going on where and where we need to raise our voice so that they are actually heard.
At the same time, we consider the transformative power of AI to be too big, broad and context -specific so that no one single institution and no single instrument will allow us to seize all opportunities and will solve all problems. So we will have to learn to live with a certain complexity of the governance of this transformation. But also, this is not a completely new situation. If we look at how we have governed the transformative power of combustion and other engines in the past 200 years, there are some lessons that we can also apply to AI. While today we are developing AI to automate cognitive labor, we have developed engines to automate physical labor. We have put engines in vehicles or machines to move goods or people from one place to another.
And we have put engines in machines to produce food or other goods automatically. And we do not expect one single institution or instrument to govern all of this. But we have developed a set of thousands of technical, legal, and also non -written societal norms that guide us in the use of these machines. We have regulated also the infrastructure that these machines use. We are setting requirements and liabilities for the people that develop, handle, and steer these machines. And we have developed instruments to protect people that are affected by the impact of these machines. And we are seeing different levels of harmonizations when it comes to regulating machines and engines. As an example, of course, we know that the airline industry is much more harmonized because it’s global than the way we regulate cars.
Cars driving in our streets on one side or the other side, where there’s more diversity possible. So after 200 years, we are still continuing to adapt the governance framework for engine driven machines, depending on the context of use. And we need to do exactly the same with AI. We need to develop appropriate technical, legal and societal frameworks and norms that allow us to develop and use AI for good in many different ways. And this work has already begun. We have analyzed our existing governance frameworks, have started to identify and fill the gaps. We have started to work on technical norms for AI systems. We have started to work on binding and non -binding legal instruments. And of course, in this regard, I’d like to particularly highlight the Vilnius Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law, for which I had the honor to lead the negotiations among 55 countries from all over the world at the Council of Europe in Strasbourg.
This provides for a principle based framework, not just for Europe, but for all countries. It provides for a principle based framework, not just for Europe, but for all countries on our planet that value human rights, democracy and the rule of law. so that our societies and economies can use AI to innovate, while at the same time we uphold our respect to human dignity and autonomy, also in the context of AI. The principles set out by the Vilnius Convention are simple and clear, but the Convention leaves enough leeway to participating states in order to allow to embed these principles in their existing legal and regulatory institutions and traditions. This will allow many countries to become parties to this global convention and to make sure that their governance frameworks may, although not become identical, but at least interoperable.
This Convention, which we hope will be ratified and enter into force very soon, will become one important instrument to make sure that AI is used for the good and not the bad. But of course, there will have to be many more binding and non -binding norms and more sector -specific norms and instruments to complement it, which hopefully will be… at least coherent in their logic and spirit. So we will use the time until the Geneva Summit next year to continue to identify gaps in global and regional governance of AI and achieve our shared objectives so that AI is used for innovation, while at the same time legitimate concerns and risks are appropriately addressed. Switzerland will be the host of the next summit, but we know that we will not be able to achieve anything on our own.
So we look forward to collaborating with all of you, with all countries and all stakeholders from the global north, south, east and west, and we will first try to identify areas where there’s a willingness and a shared vision to make progress together and then work with all of you on pragmatic and workable steps towards this vision. We will only be the facilitators trying to build bridges and build a climate of open and respectful and constructive dialogue, trying to offer pragmatic structures for trustworthy cooperation so that we can all use the potential AI, to say it again, to live together in peace, prosperity and security. Dignity. Dignity. The Swiss Summit team and I personally are looking forward to collaborating with all of you in the coming months, and we look forward to seeing you all in Geneva in 2027.
Thank you for your support and attention.
“Thomas Schneider opened the plenary by thanking the Indian government and the global audience for convening the AI Impact Summit in Delhi, and he framed the event’s three‑fold focus on “people, progress, planet” as a shared ambition for inclusive AI development.”
The knowledge base records Schneider’s opening remarks thanking India and highlighting the three sutras – people, progress and planet – as core principles of the summit [S5] and [S4].
“He emphasized that the promise of artificial intelligence must be harnessed so that everyone, regardless of geography, can share in economic and societal progress while safeguarding human dignity, personal autonomy and the health of the planet.”
Sources note that AI is seen as a tool for equitable development and that its benefits should be broadly distributed across humanity, aligning with Schneider’s statement [S45] and [S10].
“He announced that Switzerland will host the next AI Impact Summit in Geneva in 2027.”
Both the S4 and S46 entries confirm that Schneider announced Switzerland as the host of the 2027 summit in Geneva.
“He used an industry‑regulation analogy, pointing out that, unlike the highly harmonised global airline sector, car regulations remain fragmented, underscoring that AI governance will likely exhibit similar variations across domains.”
The parallel with fragmented engine (car) regulation versus more unified sectors is documented in the knowledge base discussion of AI governance analogies [S13].
“Schneider said the Geneva summit will carry a “Swiss flavour”, meaning it will be grounded in Switzerland’s tradition of constructive, neutral facilitation and will build on, rather than duplicate, existing multistakeholder mechanisms.”
S7 describes Switzerland’s approach of leveraging existing policy architectures and multistakeholder mechanisms rather than creating new institutions, matching the “Swiss flavour” description.
“Schneider then described a layered governance architecture, noting that societies have never relied on a single institution to regulate transformative technologies, illustrated with the evolution of engine regulation where “thousands of technical, legal and non‑written societal norms” now coexist.”
The knowledge base explicitly draws the same analogy between AI governance and the multifaceted regulation of engines, emphasizing the need for layered, diverse norms [S13].
Thomas Schneider consistently advocates for an inclusive, rights‑based, and environmentally conscious AI ecosystem, proposes a multi‑layered governance model that builds on existing multistakeholder platforms, highlights the Vilnius Convention as a flexible global instrument, and positions Switzerland as a neutral facilitator for the 2027 Geneva Summit.
High internal consensus – all arguments are mutually reinforcing, indicating a coherent strategic vision that can facilitate broad stakeholder alignment and practical progress on AI governance.
The transcript contains remarks only from Thomas Schneider; no other speakers are present, and all listed arguments are attributed to him. Consequently, there are no identifiable points of contention, no partial agreements, and no unexpected disagreements within the provided material. The discussion reflects a single, coherent perspective on AI governance, inclusive benefit, multi‑layered norms, and the upcoming Geneva Summit.
None – the absence of multiple speakers means the dialogue is unanimous in its goals and approaches, suggesting smooth consensus building for the topics addressed.
Thomas Schneider’s remarks shaped the discussion by moving it from a generic, aspirational framing of AI governance to a nuanced, historically informed, and pragmatically grounded roadmap. His analogies to past technological revolutions, rejection of a single‑institution solution, introduction of the Vilnius Convention, emphasis on inclusivity for less‑resourced actors, and commitment to building on existing platforms collectively redirected participants toward concrete, collaborative actions for the 2027 Geneva Summit. These pivotal comments created turning points that deepened the analysis, broadened stakeholder considerations, and set a clear agenda for future coordination.
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event
