When James Cameron wrote The Terminator, he wasn’t just creating an action movie, he was throwing a warning. A military artificial intelligence, called Skynet, became aware, took control, and decided that humans were a threat to the world, at the time, it was science fiction, but today, it seems like an urgent conversation.

Four decades after Skynet hit cinema screens, real-world AI is edging closer to autonomous warfare — and the warnings are no longer just from Hollywood.

A “terminator” From The Future. Credit: The Terminator | Indian Defence Review

In 1984, James Cameron introduced audiences to The Terminator, a film about an artificial intelligence system—Skynet—that becomes self-aware and triggers a global catastrophe. At the time, it was dismissed as dystopian fantasy. But today, as military AI evolves rapidly with limited oversight, Cameron’s warning is gaining uncomfortable relevance.

Speaking recently to CTV News, the acclaimed director reflected on his early vision, saying, “I warned you in 1984, guys, and no one listened.” His words, though cinematic, strike a chord with scientists and policy makers grappling with the rise of autonomous weapons systems—machines capable of making lethal decisions without human input.

Cameron’s concerns echo those outlined in a 2023 report by the United Nations Institute for Disarmament Research (UNIDIR), which confirmed that at least nine countries are developing or actively testing lethal autonomous weapons systems (LAWS). These include AI-powered drones, ground vehicles, and surveillance systems that could soon operate with little or no direct human control.

Libya Incident Underscores Growing Concerns

One of the earliest suspected cases of lethal autonomous action came in Libya in 2020. A report to the UN Security Council described a drone targeting retreating combatants with no confirmed human command. While the findings remain under review, they have sparked intense debate over the legal and ethical implications of such warfare.

Despite multiple attempts at regulation—primarily under the Convention on Certain Conventional Weapons (CCW)—international consensus remains elusive. Countries with advanced defense technology, including the US, China and Russia, have resisted binding agreements, citing national security and competitive disadvantage.

Cameron warns this vacuum could be dangerous. “We could be building the tools of our own destruction,” he said. It’s a sentiment shared by many in the field. In a recent report from Human Rights Watch, researchers argue that the continued development of autonomous weapons without firm regulation sets a dangerous precedent, potentially eroding humanitarian norms established since World War II.

AI Arms Race Driven by Fear of Falling Behind

Behind the push for military AI is a familiar logic: “If we don’t build it, someone else will.” This arms race mentality, Cameron points out, mirrors the Cold War, where technological escalation was fueled less by immediate need and more by fear of losing strategic ground.

Unlike nuclear weapons, however, autonomous systems are far easier to develop and deploy. They don’t require enriched uranium or intercontinental delivery systems—just data, algorithms, and hardware. That’s why analysts warn that once such technologies become widespread, controlling their use will be next to impossible.

The UN’s Secretary-General António Guterres has repeatedly called for a ban on autonomous weapons without human oversight, warning they could “take the world to a place we do not want to go.” Still, beyond non-binding resolutions, little progress has been made.

Creative AI Is Rising, but It’s Not Replacing Humans—Yet

While Cameron is vocal about the dangers of militarized AI, he’s more circumspect when it comes to its use in storytelling. The Avatar director remains skeptical about AI’s ability to truly replace writers or filmmakers. “It can replicate the structure, but not the soul,” he told CTV, noting that while AI can assist with visual effects or storyboarding, it lacks emotional depth and originality.

AI tools like Runway ML, Adobe Sensei, and Cuebric are already used in pre-visualization and editing. But winning an Oscar? That, Cameron jokes, might take another 20 years. “Let’s wait and see—if it ever wins Best Screenplay, then I’ll take it seriously.”

For now, the existential threat lies not in film studios but in covert military labs and international battlefields, where code is becoming as deadly as gunpowder—and far harder to detect.