Industry leaders push back on ‘AI bubble’ talk, say growth still grounded in real demand

Dubai: Talk of an “AI bubble” has been growing louder as global tech firms pour billions into artificial intelligence infrastructure. But, some of the world’s most influential technology leaders pushed back against that idea, saying demand for computing power and practical applications remains well ahead of supply.

NVIDIA’s CEO Jensen Huang was among the first to challenge the notion of oversupply. In a recent conversation with investor Brad Gerstner, he said the transition to AI-driven computing is far from complete. “We fully convert all general-purpose computing to accelerated computing and AI, until we do that, yes, I think the chances are extremely low,” Huang said. “Until all recommender engines are AI-based, until all content generation is AI-based, all of that’s going to be AI-generated. Until everything goes over, there’s not going to be a glut.”

Huang described NVIDIA’s production approach as driven entirely by market signals rather than speculation. “We’re at the end of the supply chain,” he said. “If they give me an order, I’ll build it. We’ve really plumbed the supply chain, from wafer starts to HBM memories, and if we need to double, we’ll double. The supply chain is ready. We’re just waiting for demand signals.”

That demand, according to other executives, remains exceptionally strong. Ahmad Alkhallafi, Managing Director for the UAE and Africa at Hewlett Packard Enterprise (HPE), said the idea of a bubble ignores how overstretched global infrastructure already is. “The infrastructure implemented for AI right now is used at 100% capacity, not able to actually cover the existing demand,” he said. “At HPE, we see supply being built to map what people believe demand is going to be, and that demand will probably outstrip that expectation.”

He added that the adoption curve is steepening. “If you look at the applications being deployed across the consumer and enterprise space, and how that’s rapidly expanding, it’s going to grow significantly in the next 12 months.”

Others say the risk lies not in overbuilding, but in misunderstanding the pace of change. Ahmad Shakora, Regional Vice President for Cloudera Middle East, said: “People worry about bubbles whenever the pace of technology innovation and hype outpaces understanding. The reality we see is more nuanced.” He pointed to Cloudera’s own findings that nearly 60% of IT leaders in the UAE have already embedded AI into their core processes, with 43% reporting significant success. “If there is a bubble, it will be one of expectations rather than of capability,” he said.

“Overpromising and under-delivering creates the most immediate harm.”

For Shakora, the path forward is discipline rather than caution. “The damage will be caused by overconfidence and rushed rollouts,” he said. “To prevent this, it is now vital to start with reliable data, tie projects to measurable outcomes, govern aggressively, and scale deliberately.”

Jeff Lunsford, CEO of Tealium, said that what some call hype is, in fact, a reflection of measurable enterprise returns. “We are not seeing a bubble or many failed projects within Tealium’s global enterprise customers,” he said. “In fact, we are seeing companies improve their conversion rates and business outcomes routinely by 25%, and even over 100% in exceptional cases.” He credited these results to “solid AI-ready customer data foundations” and the use of real-time, consented information that feeds AI systems with higher-quality signals. “Real-time streaming data provides much higher signal to AI algorithms and produces much better customer experiences across the entire technology stack,” he said.

Lunsford added that the Middle East, in particular, is approaching AI with long-term intent. “The region isn’t heading for a bubble; it’s building an ecosystem grounded in long-term strategy,” he said, pointing to national frameworks such as the UAE’s AI Strategy 2031 and Saudi Arabia’s $40 billion AI investment plan.

From a cybersecurity perspective, AI has become indispensable. Helmut Reisinger, CEO for Europe, the Middle East and Africa at Palo Alto Networks, said businesses no longer have the option of staying on the sidelines. “AI is advancing at a breathtaking pace,” he said. “It’s now woven into daily workflows, customer interactions and business strategies. We are now in a world where AI must be used to fight AI.”

He argued that the real danger lies not in over-adoption but in falling behind. “AI is not a risk in itself, it’s a capability. The real risk lies in not adopting it fast or responsibly enough,” he said. “The future of cybersecurity will depend on how effectively we harness AI to outpace increasingly sophisticated attacks, while ensuring transparency and trust.”

The consensus is that while hype can distort expectations, the fundamentals behind AI remain solid. Computing infrastructure is running at full tilt, adoption is deepening, and real business outcomes are already being recorded.