The rise of artificial intelligence has created a historic turning point, forcing nations to confront unprecedented questions about ethics, safety, and technological accountability.

As AI systems become faster, stronger, and more autonomous, the line between benefit and risk grows increasingly thin.

The panel “AI for Humanity: Ethical and Safe AI in the New Era” at VinFuture Week 2025 arrives at a critical moment  -  providing a forum for global reflection on how humanity can navigate a technology that now transcends national boundaries.

More than an academic exchange, the dialogue signals Vietnam’s entry into the global arena of AI governance. With rapid digitalization and growing ambition, Vietnam now joins the conversation shaping global standards, legal frameworks, and ethical visions for the future of technology.

Robot.jpg
Professor Toby Walsh, a leading AI expert from Australia, praised the relevance of the VinFuture panel amid intensifying global debate on AI governance. Photo: UNSW

 

AI ethics: From ‘barrier to innovation’ to foundation of sustainable progress

For years, ethics has been seen as an obstacle to technological advancement. However, Professor Toby Walsh  -  one of the world’s most respected voices in AI  -  argues that ethics is not a hindrance but a prerequisite for sustainable, safe, and predictable innovation.

His argument is grounded in reality: AI models are scaling rapidly, automating decision-making processes, and directly impacting human lives in ways even their developers cannot fully control.

When algorithms are able to analyze, assess, predict, and make decisions for humans, the lack of ethical standards becomes not just a technical flaw  -  but a social threat.

Recent incidents around the world highlight this: biased AI in hiring processes, recommendation systems that polarize society, self-driving software failing in unpredictable situations. Each is a reminder that AI ethics is not an abstract theory  -  it’s a critical requirement to maintain public trust.

In this context, the question “Should we slow down AI?” is outdated. The real issue is not speed, but direction.

AI risks know no borders  -  so global cooperation is non-negotiable

Professor Edson Prestes issued a stark warning: any AI system developed in one country can reach the world within hours via digital platforms. This reveals an uncomfortable truth  -  AI risks are transboundary, and no country can simply “close its doors” to protect itself.

There are already precedents. A news ranking algorithm creating “information bubbles” across dozens of countries. A flawed facial recognition system affecting millions globally. A leaked language model causing consequences in multiple markets.

National regulations, alone, won’t suffice.

The experts agree: international cooperation is essential. Especially as AI becomes the new infrastructure for finance, healthcare, education, defense, and governance.

The world needs shared standards  -  similar to aviation or cybersecurity  -  to prevent vulnerabilities from spreading and causing large-scale harm.

This means countries must proactively participate in shaping the future of AI, not just wait to regulate after technologies are deployed.

Vinfuture.jpg
Professor Edson Prestes emphasized that all countries must actively participate in developing ethical, responsible AI. Photo: UFRGS

Vietnam emerges as Southeast Asia’s agile voice in ethical AI

From a policy perspective, Professor Walsh’s comments on Vietnam were strategic. Though not a tech superpower like the US or China, Vietnam has unique strengths: a young population, a strong digital transformation agenda, rapid government and business adoption, a vibrant AI startup ecosystem, and openness to regulatory innovation through sandbox models.

These advantages position Vietnam as a country that, while not leading in core AI tech, can lead in building ethical frameworks and governance models  -  much like Estonia has done with digital government despite its small size.

Experts noted that Vietnam possesses three rare conditions:

A large internal demand  -  with a sizable population, aggressive public service reforms, and rapid digitalization across industries

Fast, decisive policy moves  -  from VNeID to 12 national databases and recent resolutions on AI safety

A national ambition  -  to grow the digital economy, build a domestic AI talent pool, and attract global experts
If Vietnam can define a robust ethical framework and governance model for AI, it could serve as a reference point for ASEAN  -  just as it has for the region’s digital transformation agenda.

VinFuture 2025: Vietnam finds its voice in the global AI conversation

The panel “AI for Humanity: Ethical and Safe AI in the New Era” carries more than scientific weight  -  it sends a clear message: Vietnam seeks a seat at the global table in shaping AI’s future.

With international frameworks such as the EU AI Act, OECD AI Principles, and G7 Hiroshima Guidelines now in place, Vietnam’s presence at high-level science forums is essential  -  to safeguard national interests, contribute to global standards, learn from global best practices, and strengthen regional influence.

VinFuture offers an ideal space for this  -  gathering world-class scientists in a high-quality, politically neutral environment, unlike many traditional summits.

It’s also a stage where Vietnam can share its human-centered AI approach  -  a rising trend among democratic nations and emerging economies.

This year’s panel focused on three strategic questions:

How will AI change the world?
What values must be embedded in AI to ensure fairness?
How can we develop AI that serves the common good of humanity?

These are not just the concerns of scientists, but questions for every nation facing this technological wave.

Professor Prestes delivered a sobering insight: “A nation that only uses technology will always depend on others. A nation that develops technology can protect its own values.”

This leads to two strategic recommendations for Vietnam:

First: Engage more deeply in core AI development. Competing with Big Tech is unnecessary, but Vietnam can create models, algorithms, and tailored applications to meet domestic needs  -  such as e-government, smart education, preventive healthcare, disaster risk management, and smart agriculture.

Second: Proactively build a national ethical framework for AI  -  based on transparency, data privacy, non-discrimination, system safety, and accountability.

If successful, this framework could become a reference across ASEAN  -  mirroring Vietnam’s role in regional digital policy leadership.

Distinguished speakers at the panel included:

Associate Professor César de la Fuente, University of Pennsylvania, USA – Alfred Sloan Fellow (2025), ranked in the top 1% of cited interdisciplinary researchers globally

Associate Professor Luu Anh Tuan, Nanyang Technological University, Singapore – Executive Director of the AI Research Center at VinUni, Vietnam

Professor Edson Prestes, Federal University of Rio Grande do Sul, Brazil – Head of the Phi Robotics Lab, CNPq researcher, and member of the Global Commission on Responsible Military AI

Professor Leslie Gabriel Valiant, FRS, Harvard University, USA – Turing Award winner (2010), VinFuture Prize Council member

Professor Toby Walsh, ARC Laureate Fellow and Scientia Professor at UNSW Sydney, Australia – Chief Scientist of UNSW.AI

Video presentations by:

Professor Yoshua Bengio, University of Montréal, Canada – Co-Chair of LawZero, founder of Mila Institute, UN Science Advisory Board member, and 2024 VinFuture Grand Prize laureate

Dr. Vinton Gray Cerf, Google, USA – Internet pioneer, 2022 VinFuture Grand Prize laureate

Professor Geoffrey Hinton, University of Toronto, Canada – 2024 Nobel Prize in Physics recipient and widely regarded as the “Godfather of AI,” 2024 VinFuture Grand Prize laureate

Thai Khang