DAMAC’s Sajwani sees data-center demand but warns power is hard to secure

Bloomberg’s Guy Johnson asks DAMAC Group founder Hussain Sajwani how he views recent turbulence in AI investments. Sajwani says markets fluctuate, but he believes AI will persist and change how people work and live.

Sajwani sees large financing and electricity demands ahead, but points to the commitment of major technology companies as evidence of the industry’s determination. He does not consider the market a bubble, although he expects corrections and individual failures can occur.

He cites a sharp move in a data-center company’s shares as what he considered a buying opportunity, arguing that its equipment and contracts with large customers support its business. That is his assessment of one investment, not proof that the company or sector is risk-free.

Sajwani predicts US political and environmental scrutiny will constrain data-center supply and encourage investment abroad. Nonetheless, he says the US will remain the largest market, with substantial demand.

He says his group committed $20 billion to US data centers and has bought three land parcels representing roughly 680 megawatts of load, at an estimated cost of about $9 billion. He does not expect to reach the full commitment by year-end, though he says doing so within roughly 18 months is possible if suitable land and power are available.

Sajwani describes a global data-center pipeline of roughly 6 gigawatts across the US, Europe and Asia. He says the principal constraint is finding sites with power; the figures describe his company’s stated pipeline, not necessarily operational capacity.

Asked about local concerns over power and water, he argues that data centers could pay higher electricity rates while households receive discounts. He also says newer liquid-cooled facilities reduce water use. He presents those ideas as responses to public concerns, not as a demonstrated solution for every site.

Sajwani says his group expects to acquire more land outside the US over the next 12–18 months, provided sites have power and connectivity. He cites negotiations in the UK and Australia and says expanding into new countries requires planning and logistics.

His target is a global pipeline of 10 gigawatts by around the end of the following year. The company focuses on facilities of at least 100 megawatts, mostly liquid-cooled and designed for AI chips, located near major demand centers in the US, Europe and Asia.

Sajwani says Europe and the US are more expensive to build in than Asia, citing labor and equipment choices. He estimates data-center costs rose around 30% over the preceding couple of years and could rise another 20–30%; the latter is his forecast.

As a developer, he says he expects higher costs to be passed on to hyperscale customers, while supply-chain delays are a bigger immediate constraint. He does not claim that every customer or project will necessarily absorb those costs.

He says US land with assured power can command attractive rates because demand from hyperscalers is strong. In Europe, he identifies environmental rules, zoning and lengthy regulatory processes as obstacles to the speed at which sites can be developed.

Asked about Anthropic, Sajwani says an early investment generated a gain over a much shorter period than his previous business-building career. He discusses a prospective valuation and public listing; those figures and timing are his statements during the interview, not verified completed transactions.

He recalls opening a Silicon Valley office in the late 1990s, losing money in the internet downturn and returning to technology investing years later. He says the group now has interests in AI companies and many venture funds, including Anthropic and OpenAI.

Sajwani expects other large private technology companies to consider public listings after Anthropic, but treats market pricing as uncertain. He says a small number of holdings could generate substantial gains for his group; these are his projections about investments and exits.

On finding the next opportunity, Sajwani recounts asking OpenAI’s Sam Altman about the market’s direction and being told that even a six-month outlook was difficult. His conclusion is that relationships and proximity to the market matter, but luck also plays a role.

He says the same success may not repeat and describes investing across a large portfolio, doing diligence and continually monitoring companies. He treats venture investing as a mix of preparation, diversification and uncertainty.

Johnson asks about Nvidia and its leadership. Sajwani calls it an important company and praises its chief executive’s commercial strategy and ecosystem-building, as well as its chips. He believes losing that leadership would affect the industry, while saying AI development itself would continue.

Asked whether rapid deployment should slow because of inflation and bottlenecks, Sajwani says he prefers to move quickly and address problems along the way. He cites his company’s land and power pipeline as part of that approach.

Johnson asks whether future customers will generate enough revenue to justify the investment. Sajwani believes demand will come but acknowledges that he cannot predict the timing or magnitude of a correction. He compares AI spending with categories such as mobile devices that became important to consumers over time.

Sajwani says the sector could experience significant ups and downs, but he cannot say how large. He jokes that a major correction would create a buying opportunity, then acknowledges that wishing for one would not be fair to everyone affected.