News & Views

PDF PRINT

Managing Partner, Loong Caesar recently discussed an article published by Steven Okun, Founder and CEO of Singapore-based APAC Advisors, a public affairs consultancy working with clients on ESG, sustainability and stakeholder engagement. Prior to moving to Asia in 2003, Steven served in the Clinton Administration.

Here are excerpts from their discussion on what Steven calls the AI Trilemma.

Loong: Hi Steve, its good to be back in touch with you. It may seem counter intuitive to ask in a Nvidia driven world, but should businesses be betting so heavily on AI?

Okun: AI delivers. It accelerates decision-making, drives efficiency, and opens new revenue streams, and I understand why leaders are betting heavily on it. But the same leaders are simultaneously treating AI as clean air. That is to say, always available and in whatever quantity their business needs. Building strategy and making consequential workforce decisions on that assumption is not responsible leadership. Access to AI could be constrained by cost, regulation, or geopolitics, and few are planning for that possibility.

Loong: What risks do countries like Malaysia face if AI adoption continues without guardrails?

Okun: The country could see widening inequality, higher emissions, and increased social tension as jobs are displaced. But here is why that matters commercially. Each of those downsides feeds directly back into constrained or more expensive AI access. Environmental damage from data centres drives regulatory intervention and higher utility costs. Job displacement fuels the political pressure that translates into restrictions on AI. The responsible argument is not just an ethical one. It is a business one.

Loong: Who inside companies is best positioned to help leaders think about AI risks?

Okun: Chief sustainability officers and government affairs leaders already deal with climate risk, supply chain fragility, and the social impacts of globalisation. Their experience offers a blueprint for how to future proof AI adoption.

Loong: Our law firm believes in ethical and sustainable business and you have raised material concerns. Should companies do ESG risk planning for AI?

Okun: Yes, in the same way businesses run scenario planning for cyberattacks and natural disasters, they need to do the same for AI. The parallel to climate scenario planning is instructive. Many boards already stress-test against a range of futures, from 1.5°C to 3°C+, not to predict which will materialise but to identify where the business is most fragile. The same discipline applies directly to AI. Just as there are multiple plausible climate futures, there are multiple plausible AI futures.

Loong: What is the “Responsible AI Trilemma”?

Okun: It refers to three interconnected risks:
• environmental damage: data centres consume vast amounts of energy, water, and land, and that footprint is already triggering regulatory backlash and driving utility costs higher
• job displacement
• rising inequality: AI’s productivity gains flow disproportionately to high-skilled workers in advanced economies, producing a K-shaped outcome that breeds the political instability which leads to regulatory intervention
If companies ignore these, they may face real limits on AI access.

Loong: What scenarios should companies be planning for?

Okun: There are three plausible futures:
1. AI remains abundant and affordable. This is the baseline assumption most companies operate under today.
2. AI stays available but becomes expensive. Compute costs rise, energy and water constraints tighten, and smaller firms get priced out.
3. AI becomes rationed or controlled by governments. Access is restricted for national security or economic reasons, and data localisation rules fragment the market.
Companies need to test their strategies against all three.

Loong: What happens if businesses optimise for productivity without considering continuity?

Okun: Companies that replaced workers with AI face a double hit if access becomes costly or restricted. Higher bills for compute and the loss of the institutional knowledge that walked out the door with those workers. Human expertise is not a tap you can turn back on. Rebuilding requires hiring, onboarding, and retraining an entirely new workforce, which takes years and costs considerably more than the savings ever delivered. In the most extreme scenario, a business model built around always-on, affordable AI may not be viable at all.

Loong: Are there signs that AI access might already face constraints?

Okun: Yes, data centres are being paused around the world due to energy and water constraints – from protests over power consumption in Ireland to water pollution demonstrations in Malaysia. In the US, candidates who championed data-centre development have faced electoral consequences where the local costs were most visible. These are not isolated incidents – they are early signals of a convergence that is already constraining where and how AI infrastructure can be built.

Loong: This is very true. Who could have anticipated the Hormuz blockade just 3 months ago? What measures should our clients be considering?

Okun: They need to address the environmental, social, and economic risks of AI within their own operations. Doing so reduces exposure, lowers costs, and helps defuse societal opposition.

Loong: What should businesses be asking?

Okun: Questions such as:
• What if computing costs double?
• Which parts of our business fail if we lose access to certain AI models?
• How will employees and customers react if we adopt AI without considering its societal impact?

Loong: What’s the overall message for businesses and investors?

Okun: Treat AI access as a strategic risk variable – just like carbon, water, or supply chain security. Companies that plan for multiple AI futures will be more resilient. Those that assume today’s conditions will hold may find the worst case scenario becoming the new normal.

Loong: You’ve introduced very pertinent perspectives for AI implementation. Thank you, Steve for sharing this with me.

[Steven Okun is contactable at steven.okun@apacadvisors.com ]