New survey research reveals how CEOs are navigating the next era of AI, prioritising performance over sustainability amid immense pressure to rapidly scale-up the technology, reports Sustainability Online.
The global WSJ Intelligence–NTT research poll surveyed 359 CEOs of large companies across multiple regions and sectors, transforming their results into a whitepaper that examines the state of artificial intelligence, infrastructure readiness and the changing responsibilities of technology leadership in this new era.
More than two-thirds (68%) said they planned to increase investment in AI over the next two years – nearly half by more than 11 percent or more – despite a growing share (83%) recognising the environmental cost associated with rapid AI scaling, up six percentage points from 2024.
Less than a fifth (18%) believe their technology infrastructure is optimised to support AI at scale. Meeting those infrastructure demands will only become more difficult as models become larger, workloads become heavier and expectations continue to grow, they say.
[See more: Portugal launches a national AI plan and gigafactory bid]
Performance dominates respondents’ approach to early AI infrastructure design, with seven in 10 CEOs prioritising either maximum performance or a performance-first approach to AI workloads. Three-quarters believe sustainable practices will sacrifice profitability, revealing what the whitepaper calls “a persistent zero-sum mindset.”
Where sustainable steps are being taken, efforts focus on prioritising cloud partners with documented commitments to renewable energy (56%), optimising AI models (56%) and investing in more energy-efficient hardware (54%).
While the largest share (40%) report prioritising a “pragmatic balance,” investing in the best mix of performance and efficiency currently available, less than one in five (19%) were willing to put sustainability ahead of deployment speed and even fewer (6%) planned to delay major investments until new, more efficient solutions became mainstream.
Among those solutions are green data centres and photonics, a light-based data transmission technology that enables higher throughput with lower heat and energy demands. Both are areas where Japanese technology company NTT, one of the whitepaper authors, is quick to tout its own work. Reframing sustainability as “a mechanism for strengthening performance rather than restricting it,” they argue, will put organisations in a better position to scale AI without the cost penalties associated with older, less efficient systems.
