Artificial intelligence could become a key ally for Ukrainian troops
"Vostok, I'm Sneg 02. On the highway we have to turn fucking left," says one of the soldiers in Russian, using code names that mean "East" and "Snow 02". "Received. No need to continue. Switch to defense. Over," replies another.
Later, a third soldier tries to contact a comrade, codenamed "Sud 95" : "Y ug 95, do you have contact with a superior? Warn him of the artillery fire on the highway. Artillery fire on the highway. Do not proceed in line. Move with caution".
The third soldier always seems to more agitated: "Take the radio. Tell me your situation and the position of the artillery, and approximately what weapon they are firing at." Later, he speaks again: "What's the name of your division? Yug 95, answer my questions. Tell me the name of your division!".
AI applied to military intelligence While the soldiers communicated with each other, an artificial intelligence (Ai) was listening. The military's words were automatically captured, transcribed, translated and analyzed using several AI algorithms developed by Primer, a US company that provides artificial intelligence services for intelligence analysts. It is unclear whether Ukrainian troops also managed to intercept the communication, but the increasing use of AI systems to oversee the Russian military shows the growing importance of sophisticated open source information in military conflicts.
War in everyday life: the story of a Ukrainian family from Černivci We are guests in a house in the west of the country, on the border with Romania, between beautiful wrought iron gates and luxuriant plants. And a rage that grows by the day Several unsecured Russian broadcasts have been posted online, translated and analyzed on social media. Data from other sources, such as smartphone videos and social media posts, were similarly reviewed. But it is the use of natural language processing technology for Russian military communications analysis that is new. For the Ukrainian army, the analysis of intercepted communications is still an activity that generally involves human beings, who are in charge of translating messages and interpreting commands.
The tool developed by Primer shows how the machine learning can be a valuable aid in analyzing intelligence information. The past decade has seen significant advances in AI capability in the field of image recognition, transcription, translation and language processing, thanks to neural networks that learn using large training datasets. Today, there are codes and APIs (application programming interfaces) that use artificial intelligence and are able to transcribe speeches, identify faces and perform other tasks, often with great precision. Given Russia's advantage in terms of men and artillery, intercepting communications can make a difference for Ukrainian troops in the field.
Primer sells AI algorithms trained to transcribe and translate phone calls, and others that can to extrapolate key terms or phrases. Sean Gourley, Primer's chief executive, explains that the company's engineers have made changes so that these tools perform four new functions: collect the audio picked up by web feeds that transmit intercepted communications thanks to software that mimics radio receivers; remove noise, such as background conversations and music; transcribe and translate speeches into Russian; and finally highlight the key phrases related to the situation on the battlefield. To carry out these activities in some cases it was necessary to retrain the machine learning models so that they recognized colloquial terms used to describe military vehicles or weapons.
Strategic advantage The ability to quickly train and retrain Ai models is destined to become a crucial advantage in the wars of the future, explains Gourley, adding that Prime has made the tool available to external parties, but without revealing its name: "We will not say who is using it or for what," he says. Since the start of the war, other American companies have provided technology, information and expertise to Ukraine.
The fact that some Russian troops are using unprotected radio channels has surprised military analysts, and seems to indicate a military operation with insufficient resources and training, says Peter W. Singer, a senior fellow at the New America think tank who specializes in modern wars. "Russia has used wiretapping of open communications to target its enemies in past conflicts, such as in Chechnya, and therefore, most of all, it should be aware of the risks," Singer explains. These radio signals may have helped the Ukrainians, although they were most likely manually analyzed, Singer adds. "All of this is a symptom of communications equipment failures, a certain arrogance and, perhaps, the level of despair at the highest levels of the Russian military," adds Mick Ryan, retired Australian author and general.
WiredLeaks, how to send us an anonymous report According to Harvard espionage historian Calder Walton, the invasion of Ukraine demonstrates how valuable open source information has become for intelligence workers. Facial recognition software has been used to identify people from videos of the conflict: "We have reached a watershed in the nature of information gathering and what is available," Walton adds. The ongoing conflict has highlighted how important it is to analyze different intelligence sources. It is possible, for example, that Ukrainian troops hit some Russian generals by looking in satellite images for gray-haired men near antennas, using drones or other tools. Russian troops have also started using cell phones, revealing their location and mission details in some cases, as well as their frustration and low morale.
Walton argues that most likely the National Security agency) - the main agency for the collection and analysis of electromagnetic signals in the United States - and its British equivalent, the Gchq (Government Communications Headquarters), have some versions of the tools used by Primer. The company is among the growing number of companies that could make these technologies more accessible to those working in the defense sector and private industry. The involvement of private companies in the war in Ukraine, such as those providing communications and satellite imagery, raises questions about the power these organizations have, and their role in an international conflict.
Drawing on data open source intelligence means sifting through immense amounts of information. "No one is capable of processing the enormous amount of open source intelligence information," says Emily Harding, a senior fellow at the non-profit political research Center for Strategic and International Studies, who wrote a report on the use in January 2022. artificial intelligence to obtain open source intelligence information. According to Harding, image analysis in the intelligence world has improved thanks to machine learning tools. Image content classification was one of the first fields to demonstrate the effectiveness of modern artificial intelligence. Harding adds that Primer has made a name for itself with its language analysis capabilities.
The war of the future Thanks to recent advances, AI could become a more effective tool for analyzing text and language . In recent years, artificial intelligence has developed the ability to summarize text or answer questions thanks to a particular machine learning model known as a “transformer”. These Ai models are better at understanding inputs, such as a long string of words within a sentence. Transformers have led to the creation of Ai programs that are able to generate consistent articles or use computer code to perform a certain task.
See more Choose the sportsgaming.win newsletters you want to receive and subscribe ! Weekly news and commentary on conflicts in the digital world, sustainability or gender equality. The best of innovation every day. It's our new newsletters: innovation just a click away.
Arrow However, Harding points out that the intelligence community will have to contend with the same problems that plague AI applications in other industries, such as algorithmic biases caused by unreliable or unrepresentative training data. Since machine learning algorithms often work in opaque ways, Harding adds, intelligence workers will need to find ways to ensure the reliability of the conclusions these programs come to. An incorrectly transcribed communication could have lethal consequences on the battlefield, for example putting soldiers in danger or directing a missile attack in the wrong direction.
Data collection and analysis by means of intelligence could end up playing a central role in battlefield operations. The US military is investing millions to develop AI software that can absorb and analyze various electromagnetic signals in the field. A US Army program, Tactical Intelligence Targeting Access Node, aims to create a ground station capable of obtaining information from different sensors and data sources on the battlefield. If the Russian invasion of Ukraine was largely based on centuries-old tactics such as tank maneuvers and artillery bombardments, future wars that the United States and other countries are preparing for could rely largely on new technologies. such as artificial intelligence.
Advances in technology could provide troops with critical information more quickly, allowing the military to strategize more effectively than their opponents. According to experts, however, the use of artificial intelligence on the battlefield could also trigger a dynamic similar to that of the cat and mouse, leading to a simultaneous intensification of efforts to deceive or mislead the algorithms. "Our philosophy on AI and defense is that whatever algorithm you start a war with will not be the same one you end it with," Gourley says.
This article originally appeared on sportsgaming.win US .
