Jensen Huang Wants New Social Norms for the AI Era — the Question Is Who Gets to Write Them

The CEO of Nvidia says society must adapt to AI the way it adapted to cars. He did not say who should decide what the adaptation looks like, or who pays the cost.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://open.substack.com/pub/johanosteyn/p/jensen-huang-wants-new-social-norms?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

In June 2026, Jensen Huang gave an exclusive interview to the Associated Press in Sherman, Texas, at a groundbreaking ceremony for an expansion of supplier Coherent’s manufacturing facility — a company that makes lasers used to transmit data between computer chips, cutting AI power consumption by up to 50 per cent. The interview produced a single analogy that will define the AI governance conversation for months. Cars, Huang said, were once portrayed as killing children. The world did not stop cars. It built sidewalks, crosswalks, and traffic laws. Children stopped playing in streets. Drivers learned to expect pedestrians. New norms emerged, and society adapted. “When I was growing up, I used to play in the streets,” he said. “When cars came along, you obviously can’t play in the streets now.” His argument: AI requires the same kind of societal adaptation — not resistance, but redesign. Society has no choice but to change.

The analogy is more honest than Huang perhaps intends. And that honesty is precisely what deserves examination.

CONTEXT AND BACKGROUND

Jensen Huang is not an ordinary technology executive offering an opinion on AI governance. He is the CEO of Nvidia, the company whose graphics processing units power virtually every major AI system in existence — from the models that generate language to the ones that train new AI, from the data centres that run enterprise AI to the military and intelligence systems that are beginning to deploy it autonomously. Nvidia’s market capitalisation of roughly five trillion dollars has made it the world’s most valuable company. Huang’s personal wealth is measured in the tens of billions. His interview with the Associated Press was conducted at the groundbreaking of a facility that manufactures the optical components making AI infrastructure more energy-efficient — a setting chosen, deliberately or not, to embody his argument that AI creates manufacturing jobs while reducing its own environmental footprint.

That context matters because the call for new social norms does not arrive from a neutral observer. It arrives from the person with the largest commercial stake in the outcome of the norm-setting process. When the CEO of a five-trillion-dollar company says society must adapt to AI, the governance question is not whether he is right. It is what kind of adaptation he is calling for, who gets to define it, and whose interests are centred in the definition.

Huang acknowledged that AI has become a political flashpoint — with objections to data centre construction, fears about the speed of adoption, and concerns about job losses for workers who might not have a safety net. He called for government regulation and safety standards. He said national security should be a priority. He argued that the US-China AI race is best won by a United States open to competing globally. Notably, he also pushed back against a recent political proposal — backed by figures as politically distant as Donald Trump, Bernie Sanders, and Sam Altman — suggesting that the US government should take equity stakes in major AI firms so that the public can directly share in the concentrated wealth being generated. “Americans have a stake in American companies already, naturally,” he said, citing taxes and stock portfolios. He did not propose a workers’ safety net. He did not name what the new social norms should be. He did not identify who should have a seat at the table where they are designed.

INSIGHT AND ANALYSIS

The car analogy is powerful, instructive, and incomplete — and understanding what it omits is the most important governance exercise the interview invites.

Cars did transform society. The norms that eventually made that transformation liveable — traffic laws, speed limits, road design, pedestrian protections, vehicle safety standards, insurance requirements, liability frameworks — took decades to develop. In the gap between the arrival of automobiles and the arrival of adequate governance, people died. Children died before sidewalks were built. Workers died in unregulated vehicles before safety standards were mandated. Communities were destroyed by highway routing decisions made without their participation. The eventual norms were necessary and beneficial. The people who bore the cost of the gap between the technology and the governance did not survive to benefit from the eventual settlement.

Huang’s analogy, used to argue for AI adaptation, implicitly accepts this as the price of progress. The sidewalks come eventually. The crosswalks are built.

Society adapts. What the analogy cannot accommodate is the distributional question: who bears the cost in the gap, and how does the norm-writing process ensure that those who bear the highest cost have the greatest voice in how the norms are designed?

I have previously written about the question of who sets the moral framework for AI governance — examining the Vatican’s Magnifica Humanitas as the most significant AI ethics document of 2026, precisely because it centres the question of who benefits and who bears cost in its approach to AI governance, rather than beginning from the technology’s capabilities. Huang’s interview sits at the opposite end of that spectrum. It begins from the technology’s inevitability and asks society to adapt. The Vatican document begins from the human being’s dignity and asks the technology to serve it. Both are engaged in norm-setting. Only one asks who the norms are for.

The answer to that question is not abstract. It is determined by who is in the room when the norms are written. At present, the norm-writing process for AI governance includes technology executives whose commercial interests align with rapid AI adoption, governments primarily concerned with national competitiveness, and regulators whose frameworks are perpetually behind the capability they are trying to govern. Workers who have already lost jobs to AI automation, communities whose infrastructure is being transformed by data centre construction, and countries in the developing world whose labour markets and governance institutions are most exposed to AI disruption are represented, if at all, as considerations rather than as participants.

IMPLICATIONS

For South African boards and executives, Huang’s interview raises a governance question that goes beyond the US political context in which it was delivered. The scale of what is at stake globally is illustrated by what is being debated in Washington right now. The political proposal that Huang dismissed — US government equity stakes in AI companies to ensure the public shares in the concentrated wealth being generated — is not a fringe idea. It has support from figures across the political spectrum precisely because the concentration of five-trillion-dollar valuations in the hands of a small number of companies and their shareholders is visible, measurable, and politically explosive. The Global North is actively debating state equity, national sovereignty, and massive infrastructure wealth distribution as governance responses to AI concentration. South Africa is not debating any of these questions because it currently has no functioning AI governance framework at all.

The absence of a South African policy position is not a neutral condition. While global leaders and technology executives are actively debating radical governance concepts — including state equity ownership in trillion-dollar AI firms to mitigate wealth inequality — South Africa is effectively mute on the question. South Africa’s Draft National AI Policy was withdrawn in April 2026 after citations were found to be fabrications, leaving the country’s unique structural vulnerabilities, labour market dynamics, and communities completely absent from the global norm-writing process at the exact moment the concrete is being poured. The norms that will govern AI’s impact on South African workers, communities, and institutions are being written in Washington, Brussels, and Beijing. South Africa is not in the room.

South Africa’s existing inequalities mean that the costs of the AI norm transition will not be distributed evenly. The communities and workers most exposed to AI disruption — those in roles most susceptible to automation, those with least access to the retraining and educational pathways that lead to AI-adjacent employment, those in regions where the formal economy is already fragile — are the least represented in the processes through which new AI norms are being written at every level. The pattern is consistent with every previous technological transition: the norm-writing process tends to begin in earnest after the harm has become undeniable, and the communities that experienced the harm first are rarely the architects of the remediation.

For South African organisations, the practical implication is specific. AI governance frameworks — internal policies, deployment standards, risk assessments, ethics reviews — are currently being designed primarily by technology vendors, compliance teams, and consultants whose frameworks are derived from global standards developed without significant African participation. The question of whose values, whose vulnerabilities, and whose communities are centred in those frameworks is one that South African boards should be asking explicitly before the frameworks are finalised. The workers who will be affected by the AI decisions being made in South African boardrooms right now are not in those boardrooms. The norm-writing is already underway. The question is whether it will be revisited before or after the harm it omits becomes visible.

CLOSING TAKEAWAY

Jensen Huang’s car analogy is right in its essential claim: society will adapt to AI, norms will change, and the world will look different on the other side of the transition than it does today. Where the analogy falls short is in the question it does not ask — not whether the adaptation will happen, but how it will be designed, who will design it, and what accountability exists to the people who bear the cost of the gap between the technology and the governance that eventually catches up to it.

The children who died before sidewalks were built did not benefit from the eventual norm. The workers displaced by AI before safety nets are established will not be made whole by the eventual governance framework. That is not an argument against AI. It is an argument for urgency — and for ensuring that the people with the most to gain from new AI norms are not the only ones in the room where they are written. For South Africa, the urgency is compounded by the absence: while the rest of the world debates who should own the gains of the AI era, South Africa has not yet secured its place in the conversation about who should bear the costs.

Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net