When James Cameron made The Terminator in 1984, the idea that artificial intelligence could become an existential threat to humanity belonged firmly to science fiction.
The film imagined a future in which Skynet, an advanced military artificial intelligence, becomes self-aware, concludes that humanity is a threat to its existence, and launches a devastating war against its creators. Machines eventually take over the battlefield, nuclear weapons are unleashed, and autonomous killing machines hunt the survivors.
At the time, the premise was spectacularly futuristic.
Today, however, Cameron argues that some of the underlying dangers he imagined are no longer confined to Hollywood.
In a 2023 interview with CTV News, the director was asked about the growing concerns expressed by leading AI researchers about the rapid development of artificial intelligence. His response was blunt.
“I warned you guys in 1984, and you didn’t listen,” Cameron said.
More importantly, he identified what he considers the greatest danger: the weaponization of artificial intelligence. Cameron warned that nations could enter an AI arms race comparable in some ways to the nuclear arms race of the 20th century. If one country develops increasingly powerful military AI, he argued, rival countries may feel compelled to develop their own systems simply because they cannot afford to fall behind.
That warning sounds remarkably similar to the logic at the heart of The Terminator.
But there is an important distinction.
There is no real-world Skynet secretly preparing to destroy humanity. Today's AI systems are not equivalent to the fictional superintelligence depicted in Cameron's movie.
The real concern is more complicated—and potentially more immediate.
It is about what happens when increasingly capable algorithms are connected to weapons, military intelligence, surveillance systems, cyber operations and decisions involving the use of force.
And unlike Skynet, those technologies do not need to become conscious to create serious risks.
From Science Fiction to Military Technology
The central technological idea behind The Terminator was not simply that computers could become intelligent.
It was that humans could eventually give machines enough authority to make decisions about war.
That distinction is increasingly important today.
Artificial intelligence is already being developed and deployed for military purposes. Modern armed forces use algorithms and machine-learning systems for tasks such as analyzing enormous quantities of intelligence data, identifying patterns, supporting logistics, assisting surveillance and improving battlefield decision-making.
Some applications are relatively mundane.
Others are far more consequential.
The United Nations has repeatedly warned that AI is changing the nature of warfare and raising difficult questions about human control over military decisions.
In December 2024, UN Secretary-General António Guterres told the Security Council that the rapid development of AI was creating risks that could outpace humanity's ability to govern the technology. He specifically warned about AI's potential impact on weapons systems and said that algorithms had reportedly already been used in contexts involving intelligence assessment and target selection.
That does not mean today's algorithms are independently deciding to start wars.
It means the boundary between human decision-making and machine-assisted decision-making is becoming increasingly important.
And that is precisely where Cameron's warning begins to feel less like a movie plot.
The Real Fear Isn't a Robot Army
When people hear the phrase "AI weapons," they often imagine humanoid robots carrying guns and marching through cities.
That is probably the wrong way to understand the problem.
The more significant transformation could happen without a single humanoid robot.
AI can be integrated into drones, aircraft, missile-defense systems, surveillance networks, cyber tools, military command systems and autonomous platforms.
A weapon does not have to look like a Terminator to fundamentally change warfare.
The critical question is not whether a machine resembles a human.
It is whether a machine can independently identify a target, determine what action to take, and execute that action faster than a human can meaningfully intervene.
That is where speed becomes extremely important.
Human beings have traditionally provided a natural limitation on military decision-making.
A commander receives information.
The commander evaluates it.
Orders are issued.
Subordinates interpret those orders.
The process takes time.
That time can be frustrating in conventional warfare, but it can also create opportunities for reconsideration.
A mistaken intelligence assessment can be challenged.
An order can be delayed.
A commander can ask for confirmation.
Diplomats can communicate.
Leaders can recognize that a situation is escalating and attempt to stop it.
An AI-enabled battlefield could potentially compress those decision cycles dramatically.
And if multiple countries deploy systems designed to respond automatically to one another, the danger could become less about one machine "going rogue" and more about interconnected systems escalating faster than humans can understand.
That is much closer to the concern Cameron expressed.
Cameron's Nuclear Arms Race Analogy
Cameron's comparison between AI and nuclear weapons is one of the most important parts of his warning.
He argued that nations could find themselves in an AI arms race in which every government feels compelled to develop increasingly powerful military AI because it fears that competitors will do so first.
“If we don’t build it, the other guys are for sure going to build it,” Cameron said, describing how such reasoning could drive escalation.
This is essentially a classic security dilemma.
Country A develops a new military technology because it wants to become safer.
Country B sees the development and becomes concerned that it could be used against it.
Country B responds by developing its own technology.
Country A interprets that response as evidence that it was right to accelerate its program.
The cycle continues.
Neither side necessarily wants a catastrophic war.