The burgeoning artificial intelligence (AI) revolution, heralded for its transformative potential, is concurrently casting a long shadow over global energy markets and raising profound questions about safety and security. As AI firms like Anthropic and OpenAI scramble to recruit experts in chemical weapons and biological risks, the insatiable energy demands of this technological leap are revitalising the fossil fuel industry, particularly natural gas. This paradox highlights a complex societal trade-off, where the pursuit of cutting-edge AI intersects with enduring geopolitical and environmental concerns.
In a stark demonstration of the perceived existential threats posed by advanced AI, leading US-based AI developers are actively seeking individuals with specialised knowledge in hazardous materials and weaponry. Anthropic, for instance, is reportedly looking to hire a professional with extensive experience in chemical weapons and high-yield explosives. This move is aimed at preempting the potential misuse of their sophisticated AI software, which could, in theory, be exploited to facilitate the development of dangerous agents. Similarly, OpenAI, the creator of the widely popular ChatGPT, is advertising for a researcher focused on biological and chemical risks, signalling a growing awareness within the industry of the dual-use potential of their creations. Dr. Stephanie Hare, a prominent tech researcher and co-presenter of the BBC's "AI Decoded," voiced this apprehension, questioning the safety of allowing AI systems to process information concerning sensitive chemicals and explosives, including the potential for radiological weapons.
However, this heightened security consciousness in the AI sector is occurring against a backdrop of surging energy consumption. The immense computational power required to train and operate AI models necessitates vast data centres, which are proving to be voracious consumers of electricity. This escalating demand has, somewhat counterintuitively, breathed new life into the natural gas sector. Major energy corporations such as Shell and Chevron are reportedly expanding their natural gas production and export capabilities to meet this burgeoning need. Shell, for example, forecasts a sustained increase in natural gas demand extending into the 2040s, with the fuel expected to serve as a crucial transitional energy source in Asia until around 2045.
The geopolitical implications of this energy shift are also becoming increasingly evident. The European Union, in its efforts to diversify away from Russian energy supplies following a period of significant political instability, is consequently increasing its reliance on imports from the United States. This recalibration of energy strategy means that US liquefied natural gas (LNG) is becoming a more significant component of the EU's energy mix. Projections indicate a substantial growth in available LNG export capacity over the next two years, with a global surge in LNG output anticipated by 2026. This expansion is further bolstered by developments in regions like Israel, where Leviathan production is set to significantly increase gas deliveries to Egypt by 2028, with Egypt in turn supplying neighbouring countries like Lebanon and Syria. Exploration activities in areas such as Syria's offshore and the Black Sea, involving entities like the Turkish Petroleum Corporation, also underscore the ongoing pursuit of hydrocarbon resources.
Ultimately, the current trajectory suggests that the race for global AI leadership is intrinsically linked to energy policy. Grid congestion, lengthy infrastructure development timelines, and the sheer scale of data centre energy loads are shaping how nations and corporations are approaching both technological advancement and energy security. While the AI industry grapples with the potential for misuse, the fundamental energy requirements of its progress are inadvertently reinforcing the importance of fossil fuels, creating a complex and potentially contradictory societal impact. The question remains whether the world can effectively harness the benefits of AI without exacerbating environmental challenges or creating new dependencies.