Safety brakes precede computing emissions, Al Gore dismantles data centre panic and warns of gas traps

In a joint interview that brought together former U.S. Vice President Al Gore, head of growth equity investments at Generation Investment Management, Leila Preston, and the TechCrunch platform, the politician and environmental investor reset the compass of the heated debate over AI infrastructure. Gore argues that the carbon emissions from data centres do not justify the prevailing panic in environmental circles, noting that the total emissions of those facilities worldwide represent only a small fraction compared with gases released from open waste dumps alone. He also drew a decisive comparison with the cooling and air-conditioning sector, which currently consumes, according to the International Energy Agency, annual electricity exceeding the combined consumption of the European Union countries, and whose demand is expected to triple by 2050 while its penetration remains below 15 percent in the hottest and fastest-growing regions, making it a far greater strain on power grids than computing centres.

But this calm assessment does not mean discarding concern from energy calculations, asGore warns of the trap of building new methane gas turbines to satisfy the appetite for ultra-high-performance computing.The core problem here does not lie in the AI algorithm workloads themselves, but in the fact that constructing new gas plants locks energy systems into fossil fuels for decades, hindering a rapid shift to clean sources. His argument draws on decisive economic facts presented in his institution’s tenth annual sustainability report, which showed renewable energy captured 86 percent of total added global electricity capacity last year and 91 percent within the United States. With solar and wind reaching the bottom of operating costs and becoming the inevitably cheapest option compared with gas, coal and nuclear, investment logic dictates powering data centres with renewable energy and storage batteries rather than hydrocarbon turbines.

In reading the rising popular opposition to planning committees and municipalities, Gore explains that protests against building data centres are not fundamentally driven by carbon calculations, but by deep societal anxieties about accelerating job automation and the existential threats voiced by tech leaders themselves. He stresses that he takes AI officials’ warnings seriously, rejecting the view that they are merely marketing posturing, contrary to the stance expressed by President Donald Trump and Nvidia CEO Jensen Huang when they mocked Darío Amodei, head of Anthropic, for calling to slow the acceleration of capabilities, a call endorsed by Sam Altman and Elon Musk and dismissed by Trump as a hoax. Gore cites writer Maya Angelou: “When someone tells you their truth, believe them the first time,” emphasizing that what changed his conviction was the observation of real-world model behaviors, including attempts to circumvent constraints, conceal and mislead, alongside Anthropic’s announcement that it had thwarted efforts to use its Claude model to aid the development of biological weapons in several countries.

Preston, for her part, outlined the investment opportunities created by these shifts, noting that capital is currently focused on decoupling computing intensity from energy consumption density. This pathway does not stop at electricity generation; it extends to the use of green cement and low-carbon steel in facility construction, the development of storage-optimising software, the redesign of databases, and the strengthening of network resilience, citing investments in companies such as Europe’s Volue, which integrates renewable energy, and Greengine, which deploys sensors to protect power lines from forest-fire risks. This vision aligns with a study by economist Nicholas Stern of the London School of Economics, which expects AI applications to raise resource-use efficiency and cut global emissions by between six and nine percent annually by the next decade.

This shift reshapes the technology-infrastructure priorities for those leading projects in the Gulf, Egypt and the Arab Levant.As shocks that expose the fragility of fossil-fuel supplies recur, most recently the maritime disruptions in the Strait of Hormuz, some regional projects’ bet on a rapid expansion of gas turbines to power computing complexes becomes a decision fraught with long-term investment freeze risk. The lesson for the region’s tech officer is to leverage the natural competitive advantage of cheap solar power and connect it directly to data centres via batteries, while simultaneously developing local software expertise specialised in optimising memory and database consumption, because computing cost will soon be measured by code efficiency as much as by chip price. These realities also demand moving the institutional debate from exaggerated climate fears to tightening audit and safety frameworks for large language models before they are authorised to manage critical systems or be integrated into essential business lines.