In electronic warfare, AI could change the game

Over the past few decades, war has followed the same pattern as everything else, by becoming faster and more complex. A look back in time makes this point clearly enough. At the outbreak of the First World War in 1914, military communications relied largely on wired telegraphy and dispatch riders, with only a few dozen radio transmitters and receivers across the entire French army. A century later, the contrast is stark. Every soldier now carries at least one communications device as part of their kit. In today’s conflict, two armies facing each other along a front stretching hundreds, even thousands, of kilometres generate a huge volume of information that has to be processed in real time to secure a tactical edge.

Within a matter of decades, electromagnetic warfare (EW) has become a fundamental capability for modern militaries. By detecting, interpreting, controlling, or disrupting signals across the electromagnetic spectrum—such as radar, radio, or infrared transmissions—armed forces can defend against threats and anticipate enemy movements.

“Electronic warfare⁽¹⁾ is about analysing the adversary in front of you in order to obtain actionable intelligence, with the tactical goal of either protecting yourself or neutralising the enemy. Unlike other approaches, the objective is not to be exhaustive: a degree of uncertainty is acceptable, because speed is critical,” explains Frantz Loutrel, Director of Architectures and Solutions for EW and Intelligence at Thales.

^{(1)This is what distinguishes electronic warfare (EW) from signals intelligence (SIGINT), whose aim is to gather highly precise information without immediate urgency, in preparation for future action.}

Reducing mental load and making sense of complexity

In combat, agility can be a matter of survival, with threats needing to be detected and neutralised as quickly as possible. Yet soldiers in charge of electronic warfare may be operating under intense pressure and dealing with fatigue, as well as managing basic needs such as eating, finding cover and getting rest. “For all these reasons, soldiers in the field are quite logically not operating at 100% of their capabilities all the time,” says Frantz Loutrel. 

This is where artificial intelligence can make a genuine difference: By easing the operator’s mental workload and helping them make better sense of complex situations over the course of an engagement.

One striking example developed by Thales is the AI-enhanced Com tool, which can identify communications signals across different frequency bands. Com is able to isolate signals of interest, recommend appropriate countermeasures, identify similarities between signals and highlight the parts of a signal that contain the most useful information. All of which reduces the operator’s cognitive burden and makes the job more manageable.

AI can make systems more adaptable over time

Com is not just effective at detection; it can also evolve over time. As François Sausset, Data Science Team Leader and co-author of the Thales position paper AI in Electromagnetic Warfare, explains: “If AI has detected and identified an unknown source, that information can be integrated into the system rapidly, making it more effective. This was not nearly as quick in the past. It is crucial in a context where the tempo of war is accelerating, and communications tools and frequencies are constantly changing.”

Thales is also incorporating AI into an airborne audio data analysis solution (KIA). This works both at the level of the audio extract — improving signal quality, isolating voices and background noise, transcribing and separating speakers — and at database level, by detecting speech activity and grouping recordings according to a range of criteria in order to cross-reference information.

“This solution makes it much easier to process the heavily degraded signals that often come from intercepted communications,” confirms François Sausset. 

“The tests we carried out with operational personnel showed that our tools helped them to work more effectively, stay more focused, and feel less fatigued at the end of a mission.”

Using AI to learn from past operations

Helping operators react quickly, sort through large volumes of information, and free up cognitive “bandwidth” are just some of the promises offered by AI in electromagnetic warfare. Com and KIA are already a reality and are currently being integrated into Thales solutions.

Artificial intelligence can also play a major role in the feedback process. “An adversary, especially one with significant military manpower, operates according to a doctrine and tactical patterns its units have been trained to follow, and which it has already used in previous operations or conflicts,” says Frantz Loutrel. “AI is capable of retaining all this historical data effectively. It can then provide relevant feedback based on that information, helping to anticipate the adversary’s next decisions more accurately.”

Whether it is embedded into detection or jamming systems, or used to support operational feedback, artificial intelligence will not replace human intelligence sources, nor other forms of technical intelligence such as imagery or cyber. Its large-scale adoption in military equipment will depend on the practical value it delivers day in, day out to operators in the field. 

But one thing is clear: with the return of high-intensity conflict, and with the number and complexity of emitted signals continuing to grow, AI is likely to be one of the tools that could make a real difference in the decades ahead. 

“The key issue with this technology will be making sure we do not become dependent on others, and that we give ourselves the means to remain sovereign,” Frantz Loutrel stresses, “which means equipping ourselves with AI systems that are consistent with our ethics and our values.”