AI for Good: why ethics and accountability must keep pace with innovation

AI for Good: why ethics and accountability must keep pace with innovation

key takeaways.

  • AI's long-term impact will depend not only on technological progress, but on AI ethics, governance, transparency, and accountability evolving alongside increasingly capable AI systems
  • At an AI for Good event hosted by Lombard Odier in partnership with TIGER 21 Switzerland, speakers agreed that responsible AI requires explainability, human oversight, and clear ownership, rather than waiting for regulation to catch up
  • For AI companies, lasting success will depend on more than technical capability. Deep domain expertise, explainable AI, and responsible deployment are becoming key competitive advantages
  • From cybersecurity to financial inclusion, AI founders showed how AI is already tackling real-world challenges, while reinforcing the importance of human oversight, accountability, and trust.

As artificial intelligence reshapes economies, industries, and societies, the defining question is no longer simply what the technology can do, but how it should be used.

That was the central theme of an evening hosted by Lombard Odier in partnership with TIGER 21 Switzerland, ahead of the United Nations' AI for Good Global Summit in Geneva. Bringing together investors, entrepreneurs, and technology leaders, the event explored the opportunities and responsibilities emerging as AI becomes more powerful and pervasive.

Opening the evening, Duncan MacIntyre, Partner and UK Region Head at Lombard Odier, framed AI as a catalyst for profound economic and societal transformation, while arguing that "the most interesting question is not what AI can do, it is what we choose to do with AI." Florian Kemmerich, Chair of TIGER 21 Switzerland, echoed that sentiment, inviting participants to view AI as an opportunity "to learn, to engage, to understand" and ultimately as "a real blessing" for humanity.

What followed examined that question through three complementary lenses: the technological forces driving AI's rapid evolution, the governance and investment frameworks needed to steer it responsibly, and the founders already confronting these challenges as they build the next generation of AI businesses.

The most interesting question is not what AI can do, it is what we choose to do with AI

The tenth machine age

Stephen Ibaraki, founder of the AI for Good Forum, opened proceedings with a sweeping account of where he believes technology is heading. He built his first AI-powered computer at the age of ten, already driven by a determination to one day put it to good use – instilled by strong family values around doing good and giving back. That instinct carried him from advising Microsoft in its early years to counselling a coalition of large financial institutions on innovative spending and technology strategy.

His central thesis is what he calls the "S11 shift": eleven converging technologies – from nuclear fusion and quantum computing to advanced robotics and biotechnology – organised into a self-reinforcing flywheel of energy, hardware, fabric, intelligence, and autonomy. Stephen Ibaraki argues this convergence is ushering in a "tenth machine age," which he believes could unlock multi-trillion-dollar economic opportunities over the coming decade. Among the risks he flagged was the growing threat quantum computing poses to encryption, a concern serious enough that the banks he advises have asked him to return year after year for updates.

Yet the opportunity, he stressed, comes bound up with transformative risk – geopolitical, economic, ethical, and social – and with it, a need for accountability, human oversight, and transparency as these technologies scale. Staying ahead of that change, he suggested, requires a willingness to question assumptions, experiment continuously, connect ideas across disciplines, and build diverse networks.

A mirror to humanity

Michael P. Nash, director of an upcoming documentary series executive produced by Leonardo DiCaprio, followed with a preview ahead of the film's premiere. He explained that the project began two years ago with a single, open question: can AI help solve the world's greatest problems, and can it help build a future we look forward to. Most people he interviewed, he found, sat somewhere between the two extremes dominating the public conversation – technology as existential threat versus unqualified promise – hopeful, yet uneasy about the world their children and grandchildren will inherit.

Technology alone will not determine the outcome. This has to be intentional. AI will eventually hold a giant mirror to humanity

One conversation reshaped the film's direction entirely: an indigenous elder from Maui redirected his questions away from AI and towards what she called the "most important" intelligence of all – ancestral intelligence. It became a throughline of the documentary, voiced in the screened preview through the perspective of Mother Earth: "I ask you to rethink your future." For Michael P. Nash, the message was clear – technology alone will not determine the outcome. "This has to be intentional," he told the room, arguing that AI will eventually "hold a giant mirror to humanity" – and it is up to us, today, to decide what that reflection shows.

Investing, innovation, and global impact

That call for intentionality carried straight into the panel that followed. Moderated by Stephen Ibaraki, it brought together four experts – spanning investing, entrepreneurship, policy, and youth impact – each contributing a different lens on what it takes to build AI for good.

If you don't focus AI on improving humanity and our society, it can escape us

Florence Kiss, Impact expert at Lombard Odier, set the tone early, cautioning against taking AI's sustainability claims at face value. Drawing on Lombard Odier research, she observed that headline efficiency gains – carbon savings from AI-driven optimisation, for instance – can be easily offset once the full picture is considered, from the energy and water intensity of data centres towaste and impacts across the value chain. Governance, she argued, needs to be reinvented for the AI era precisely because the technology is still maturing. Her verdict was blunt: "if you don't focus AI on improving humanity and our society, it can escape us."

Read also: Why a changing world is strengthening the case for sustainable investing – from system stress to structural opportunity

Agentic AI and the shift to orchestration

If governance defines the conditions for responsible AI, commercial success depends on a different set of capabilities. For Ben Robinson, Co-founder and CEO of Aperture, this success increasingly hinges on orchestration. Bringing a fintech investor's lens, he described the industry as moving through a "phase change" – from early point solutions focussed on cost savings toward agentic AI capable of running entire workflows. That shift, he said, raises the bar considerably: it demands deep domain expertise, proprietary data, and,in regulated sectors like financial services, a deliberate approach to explainability from day one.

We should not wait for liability to be defined, but define it ourselves

Governance also means acting before regulation catches up. Drawing on his background in policy and regulation, Hannes Chopra, CEO of RaSa Future Fund and RaSa Consulting, offered a principle for boards: "we should not wait for liability to be defined, but define it ourselves." In practice, he argued, that means every AI systems interacting with customers or employees needs a clearly accountable owner and a defined process before deployment – paired, counterintuitively, with a genuinely open environment where employees at every level can experiment safely with the technology.

Either we will see an accentuation of inequality, or there will be a chance to level the playing field

Closing the AI opportunity gap

Taha Bawa, Co-founder and CEO of Goodwall, brought a Gen Z perspective, focussed on access. Goodwall's "learning to earning" pathways aim to enable young people worldwide to use AI to build small businesses and explore new income-generating opportunities. In his view, the stakes are structural: "either we will see an accentuation of inequality, or there will be a chance to level the playing field" – and realising the latter means bringing communities along, not just deploying the technology.

Closing the session, Stephen Ibaraki asked each panellist what they would refuse to optimise with AI, even where the economics were attractive. The answers converged on a shared theme: keeping humans accountable. "We have to keep humans in the equation. We have to keep accountability," said Florence Kiss. Hannes Chopra was equally direct: "we should never trust an AI system more than the people responsible for it allow us to," adding that trust in human relationships – from hiring decisions to partnerships – should never be outsourced. Ben Robinson put it most succinctly: "I don't think we can delegate ethics to AI."

We have to keep humans in the equation. We have to keep accountability

AI in practice: founders tackling real-world challenges

The evening closed with a Dragon's Den-style showcase, where founders pitched a panel of experts on ventures spanning governance, sustainability, finance, cybersecurity, and inclusion. Beneath the range of applications, a handful of shared challenges emerged.

Keeping humans in the loop

Maintaining meaningful human oversight emerged as one of the founders' most immediate commercial challenges. As AI agents take on more autonomous, end-to-end decision-making, a number of founders noted that supervision hasn't kept pace. One founder argued that organisations are moving quickly to deploy agents capable of executing complex tasks with limited human review. Their response: building systems that make each step of an agent's decision-making visible and auditable, rather than relying solely on the final outcome.

Fighting AI with AI

Security offered perhaps the clearest example of AI's accelerating arms race. As criminals increasingly use AI themselves – to mine leaked data for attack paths, or exploit the anonymity of digital assets – several ventures pitched AI as a countermeasure, built to find exposures before bad actors do and bring investigative rigour to a fast-growing on-chain economy.

Read also: Exploring ‘Silicon Valhalla’ – can AI be built in Europe?

Unlocking access with AI

Beyond efficiency, several founders maintained that AI's greatest potential lies in expanding participation – framing its value less as automation than as a bridge into markets that have historically excluded people. Examples ranged from building financial identities for women without access to credit, to helping investors assess asset classes such as natural capital more efficiently than traditional processes have allowed.

Yet for all the technological ambition on display, the founders repeatedly returned to a familiar conclusion: AI alone is not enough. The harder, more durable work lies in regulatory navigation, trust-building, and translating AI's outputs into something a human is willing to act on. As several participants observed, AI can widen access and sharpen judgement, but it cannot substitute for the accountability and trust that still have to be earned.

The defining challenge is no longer whether AI is capable, but whether it can be deployed responsibly

The human responsibility behind AI

The event showcased no shortage of ambitious ideas, from quantum computing and autonomous agents to new models for finance, cybersecurity, and inclusion. Yet beneath the technological optimism lay a striking consensus. The defining challenge is no longer whether AI is capable, but whether it can be deployed responsibly. As the technology becomes more powerful and pervasive, accountability emerged as a constraint on innovation, but as the foundation that will determine whether AI ultimately earns society's trust.

important information

This is a marketing communication issued by Bank Lombard Odier & Co Ltd (hereinafter “Lombard Odier”).
It is not intended for distribution, publication, or use in any jurisdiction where such distribution, publication, or use would be unlawful, nor is it aimed at any person or entity to whom it would be unlawful to address such a marketing communication.

Read more.

get in touch.

Please select a category

Please enter your firstname.

Please enter your lastname.

Please enter a valid email adress.

Please enter a valid phone number.

Please select a country

Please select a banker

Please enter a message.


Something happened, message not sent.
let's talk.
share.
newsletter.