Amid anxieties over American overstatement and fears of Asian rise, artificial intelligence is being enlisted in a battle that could redraw the global map of power. Technology has long been employed in warfare, but superartificial intelligence—or what is called “superintelligence”—will be an entirely different entity.
Some observers point out that we are heading toward a world where artificial intelligence systems are built to serve specific national goals, with advanced technologies turning into “pawns”—or perhaps “chess engines”—in a conflict of competing objectives among nations. One might ask whether we should reject and outrage this trend, or if it is an unavoidable reality?
Experts note that few people will care about governance issues and artificial intelligence policies unless there is a real and imminent danger; such is the nature of the human mind. Yet before delving into discussions about ethics and strategies, we must ask: how close are we to the emergence of an artificial intelligence that deserves the name “intelligence” in the human sense?
Heads of advanced artificial intelligence laboratories currently believe that a model smarter than any Nobel laureate across multiple fields such as biology, programming, mathematics, and engineering could become available between 2026 and 2027—that is, within two to three years, according to statements by Dario Amodei, CEO of Anthropic. Recent surveys conducted by research entities with 2,000 artificial intelligence researchers indicate a 10% probability of the emergence of “human-level artificial intelligence” by 2027, representing a major leap in expectations compared to previous years.
Read also: Artificial intelligence and dictatorships: Opportunity or threat?
Beyond the debate over timelines, specialists emphasize that we are approaching the automation of human intellect—from genius embodied in a silicon chip, to systems that may outperform individuals or even groups in analysis and decision-making. As for the next step, sometimes referred to as “superintelligence,” it may launch within the next decade, with leaders in prominent companies believing that creating an artificial intelligence surpassing overall human capabilities could be achieved within a timeframe not exceeding 15 years.
If these estimates are correct, can artificial intelligence remain “cross-border” without nations entering a race to control it? Initially, companies like Meta seemed open to making their models available to everyone, including open-source conversational chat models, but policies are gradually shifting as concerns grow over the deployment of artificial intelligence to reinforce the dominance of certain countries or regimes.
Anthropic CEO Dario Amodei recently stated that the future of the democratic world depends on having an “upper hand” in artificial intelligence, because the likelihood of the rise of an “authoritarian artificial intelligence” looks like an unimaginable nightmare. Similarly, Sam Altman, CEO of OpenAI, spoke about the necessity of maintaining U.S. leadership in this field, out of fear that other countries with more rigid systems might pull ahead. This has driven some tech giants—such as Microsoft and Amazon—to cooperate with U.S. defense and intelligence agencies in developing AI-enhanced applications.
In a recent memo issued by the White House, the phrase was stated clearly: “Harnessing artificial intelligence to achieve national security goals,” accompanied by mentions of China and other competitors who may dedicate massive budgets to catch up with and surpass advanced artificial intelligence.
Read also: Armies race to replace human soldiers with artificial intelligence robots.. Experts warn of disaster
On the other hand, we already find actual examples of artificial intelligence deployment in warfare, such as Russia and Ukraine using semi-autonomous drones, and military entities employing AI-driven targeting systems with limited human intervention. True, the military employment of technology is not new, but superintelligence, if realized, could represent a revolution threatening to completely change the rules of military engagement and the balance of power.
Is the dream of a global society where everyone works together to regulate artificial intelligence—akin to CERN’s nuclear physics project—merely an illusion? Some may see it as possible, or at least worth trying for the benefit of future generations. Yet the signs of ongoing “localization” or “nationalization” of artificial intelligence suggest that we are heading toward an era where the world will witness the formation of economic and military alliances centered around algorithmic technical expertise.
Even if we refuse to accept this outcome, the features of the international race for artificial intelligence are clear: procedures and legislation aimed at protecting local models, pumping massive government investments into research, and establishing partnerships between technology laboratories and security and defense agencies.
Against this backdrop, the question of ethics and international coordination among major powers remains wide open. Will we witness the fueling of a new Cold War, but this time in a digital arena? Or will there be sufficient collective awareness to formulate regulations governing these technologies before they outpace our ability to control them?
