The Artificial Intelligence Era Faces a Threat from Directed Energy Weapons

The Emerging Artificial Intelligence Era Faces a Growing Threat from Directed Energy Weapons

Autonomous and AI-enabled systems increasingly rely on optical and radio frequency sensors and significant computer power. They face growing vulnerabilities from directed-energy laser and microwave weapons

By David C. Stoudt

The U.S. Navy San Antonio-class amphibious transport dock ship USS Portland conducts a Solid State Laser – Technology Maturation Laser Weapon System Demonstrator test May 16, 2020 in the Pacific Ocean. The ship successfully knocked down a small drone using its new laser directed energy weapon in the test.

U.S. Navy/Alamy Live News

In May the U.S. secretary of the Air Force flew in an F-16 that engaged in a mock dogfight over the California desert while controlled by artificial intelligence. Carmakers from San Francisco to Boston are jousting to deliver driverless cars. In Norway a crewless cargo ship carries fertilizer from port to port. On the land, sea and in the air, we face the coming of such autonomous platforms—some envisioned to benefit humanity, and others meant for destruction—available to everyone, to governments, businesses and criminals.

These platforms possess a little-understood Achilles’ heel in the form of another fast-moving technology: “directed energy” laser and microwave weapons. The public needs to understand this weakness as we move into an era of thinking machines.

Directed-energy weapons (DEWs) can disrupt or damage the many sensors and electronics that such autonomous systems depend on. They appreciably expand the range of effects typically generated by guns and missiles, as well as conventional “electronic warfare” systems that jam or spoof communications receivers, radars and missile seekers.


On supporting science journalism

If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


That matters because the innovative sensors that utilize optical and radio frequency bands of the spectrum are driving the development of these autonomous machines, alongside artificial intelligence processing that seeks to mimic human thought or reasoning. The era won’t merely restrict itself to one of “Level 5” fully driverless automobiles or drones delivering goods and services. We will also see new weapons emerge.

Since the 1960s the U.S. government has made significant investments in DEWs, which create an intense electromagnetic environment to disrupt, damage or destroy targets in combat. While DEWs have been popularized in science fiction for over a century, they are now becoming an operational reality.

In recent years the Department of Defense has been investing more than $1 billion per year in DEW technologies that typically take the form of high-energy lasers or high-power microwave systems. Some laser and microwave weapons have already been deployed in the field while others are in testing. Laser weapon systems can typically defeat targets at a greater range than microwave systems, but a single laser can engage with only one target at a time. Microwaves, instead, given their much broader area coverage, can engage multiple targets simultaneously, if they are within the effective range of the weapon. Today’s microwave weapons are now being employed to defeat swarms of armed drones, causing them to fall from the sky by overpowering and disrupting their internal electronics. Such drones continue to be used heavily by both sides in Ukraine’s fight against Russian invading forces. And recently the Department of Defense stood up the Replicator Initiative to accelerate autonomous systems to warfighters, which may be vulnerable to DEWs.

Most laser systems fire high-power continuous waves, which means they must stay aimed at a specific part of a target long enough to damage it or its electronics. The kinetic effect that laser weapons have on targets, much like a blowtorch at a distance, is well understood and well-funded by military leadership in multiple countries. Autonomous systems’ optical sensors are especially vulnerable to laser attacks. That includes strikes from commercially available high-power lasers. These commercial lasers, and even lower-power lasers, can be combined with rudimentary optics to fashion potentially effective counter-sensor weapons.

Microwave DEWs, often referred to as radiofrequency weapons, are meant to damage electronic systems. Most in development today rely on the creation of extremely high peak powers, in the range of one million Watts to one trillion watts of effective radiated power, with pulse durations of typically less than one millionth of a second, that disrupt, degrade or damage a target’s electronics. Potential targets include not only autonomous systems, but all modern electronics. That includes everything from the electronic controls of the electric power grid to the electronics that power the entire Internet of Things. Radiofrequency weapons either couple energy into a target’s intended electromagnetic apertures, known as “front-door” coupling, or through unintended ones, called “back-door” coupling. One example of front-door coupling is attacking through the antenna of the targeted receiver. In back-door coupling, energy connects with circuits within the targeted system via seams, nonconductive surfaces or unshielded wires, creating transient voltages that can disrupt operation or even cause electrical arcing and damage within interior microchips. Even optical sensors can be susceptible to microwave attack since their response is quickly converted into electrical signals. Unlike conventional electronic warfare, these effects often continue in the electronics after the radiofrequency illumination has ended. These long-lasting effects make drones fall when hit with microwave energy.

Real-world artificial intelligence is extremely challenging, and there is much debate on whether or not such systems will ever be able to fully understand and operate in real world scenarios. As autonomous vehicle manufacturers continue to pursue driverless capabilities, they often employ different types of sensors to achieve their objectives. For example, Tesla has recently decided to only use optical cameras for its Tesla Vision–enabled cars, while Alphabet’s Waymo vehicles use a combination of radar, lidar (light detection and ranging) and cameras. Similar debates occur around the various levels of drone autonomy and the required associated sensors. Amid these discussions, we should not overlook the fact that all of these platforms also contain significant computer processing capabilities, often referred to as edge computing, that also need protection. Whatever the outcome of the debates, it is important to keep in mind that technology developers are focused on pushing the state of the art. Regardless of the system, developers are not very likely looking at future electromagnetic-environment trends. The time to consider these, however, is now, while systems are being designed, so lower-cost measures can enhance their survivability. Trying after the capability is already in operation is a much more expensive and challenging task, one that often fails.

Present-day DEWs have already proven effective, and the U.S. is one of many countries conducting research in this field. In addition, the perceived lack of implementation of hardening techniques against microwaves and lasers may drive accelerated investment in worldwide DEWs by all stakeholders (governments and bad actors alike).

As Nikola Tesla wrote in My Inventions, first published in 1919 in the Electrical Experimenter magazine: “Telautomata will be ultimately produced, capable of acting as if [possessed] of their own intelligence, and their advent will create a revolution.” Here he speaks to potentially DEW-vulnerable autonomous systems that possess artificial intelligence capabilities. Tesla’s revolution is now upon us.

This is an opinion and analysis article, and the views expressed by the author or authors are not necessarily those of Scientific American.

Read More

David C. Stoudt