Throughout military history, warfare has been transformed by innovations such as gunpowder, tanks, aircraft, and nuclear weapons. Today, we are witnessing a transformative shift dominated by the concept of autonomy powered by Artificial Intelligence (AI) in weapons systems. AI-based systems can impact all levels of warfare, from intelligence gathering, surveillance, and reconnaissance to decision-making through Lethal Autonomous Weapons (LAWs). LAWS can engage targets based on pre-determined algorithms, thereby reducing the need for direct human input in the decision to use force.
Over the past decade, militaries around the world have invested heavily in AI-based weapons platforms, such as drone swarms and unmanned sea, land, and air autonomous vehicles. The “Artificial Intelligence in Military Global Market Report 2024” predicts that the global AI in military market will grow from $8.45 billion in 2023 to $9.86 billion in 2024 at a compound annual growth rate of 16.6%. This increase in military spending indicates a large increase in the acquisition of AI-based weapon systems. It also reflects a global race to dominate this technology. As a result, AI has become an integral part of the military strategies of major and medium-sized powers.
The fusion between political digital infrastructure and warfare, between technology companies and armaments, between algorithms and artificial intelligence will be the main features of 21st century wars. Supposedly harmless algorithms for targeted advertising are already being used today to identify supposedly legitimate military targets, with significant “collateral damage”. Highlights: Following the October 7 attacks by Hamas, the Israeli military used an artificial intelligence system called Lavender in Gaza, among many others, which does not display ads but automatically puts people on a kill list once data patterns collected through surveillance match the data patterns of the alleged fighters.
The use of Lavender and other AI-powered systems in Israel’s war on Gaza has introduced an unprecedented glimpse into the shadowy world of algorithmic warfare. In Israel’s case, the basis for this is data collected over many years from mass surveillance of the Palestinian population. AI systems like Lavender were then merged with data from alleged known Hamas members and then scored almost every person in Gaza on the likelihood of being a supporter of the militant group. Up to 35,000 identities were identified as terrorists and presumed legitimate targets. The human check essentially consisted of briefly listening to audio files associated with these identities to confirm that the person in question was male. If this was the case, an airstrike was launched on the person in question, preferably in the victim’s private apartment. In the case of simple Hamas suspects, who were identified only on the basis of their scores, up to twenty civilian casualties were considered tolerable. The Israeli military assumed a ten percent error rate, which was also described as tolerable.
For data collection, the Israeli military relied on commercial cloud providers, especially Microsoft, Google, and Amazon. According to reliable sources, the cooperation with Amazon Web Services (AWS) was and is particularly close. The private providers have two things to offer: An almost inexhaustible amount of computing and storage capacity on the one hand, and advanced AI applications that are supposedly superior even to the Israeli military’s AI applications. In 2021, Israel signed a contract with Google and Amazon for $1.2 billion, which subsequently set up data centers in Israel at lightning speed, in which almost all data on the Palestinian population has since been stored. It is noted that the merger of the military, armaments and IT industry is not a new process or one limited to Israel.
Cyber-centric or digital warfare is changing the way we know war. Technology companies will play a comparable role in the 21st century to Lockheed Martin’s role in the 20th, Google CEO Eric Schmidt predicted in 2013. That same year, Amazon Web Services was awarded a $600 million contract to provide cloud services to the CIA. In 2020, this contract was replaced by the Commercial Cloud Enterprise project, which also included other major companies in addition to Amazon, such as Google, Microsoft, Oracle and IBM. However, in addition to the well-known big IT players, start-ups are increasingly entering the market and have been included, for example, in “Project Maven”, in which the US Pentagon has developed technologies that can be used to evaluate drone video on a large scale using AI. There have also been similar projects for voice recognition, facial and gesture recognition from surveillance cameras, and merging disparate data sources powered by AI to profile and track individuals or networks. Companies like Palantir and entrepreneurs like Elon Musk boast about their central role in the Ukraine war, side by side with top military officials about AI-powered warfare in the 21st century.
Artificial intelligence, in simple terms, can be defined as a computing system that can perform cognitive functions while simultaneously adapting and improving itself to become more effective. AI emulates human intelligence algorithmically, using symbols and high-level logical rules, exponentially replicating human intelligence and having the ability to make autonomous decisions. The science of AI is technologically on par with the fission of the atom and nuclear weapons.
But AI is not like a nuclear weapon. It is not a big tangible thing that can be easily detected, tracked, or interdicted. It is a science, like physics or mathematics. Applications of AI will lead not only to incremental improvements in the capability of weapons systems, but will require a fundamental recalculation of what constitutes deterrence and military power. For example, combining AI elements such as visual recognition, language analysis, automated extraction of local hierarchies (or ontologies), reinforcement learning-based systems control, simulation-based prediction, and advanced forms of search with existing technologies and platforms can quickly yield entirely new and unforeseen capabilities. Integrating AI into an existing platform is a surprise in itself. But the complex interactions of such platforms with others like them can create exponential, insurmountable surprise. What current conventional defense system prevents such an AI creation? None.
Artificial Intelligence will lead to the most significant capabilities and technologies humanity has ever built, and failure to understand the nature of AI will set us back in terms of harnessing all it has to offer in the short, medium and long term. The stakes are high beyond description.
