Defence and Artificial Intelligence. Where does Europe stand?

Key points:

• The US and China are the main players when it comes to AI and its military-related development, with Russia also aiming to be in the AI race.
• Big Tech companies have become key geopolitical players, often shadowing national governments.
• Europe's investment has increased significantly during the last years, but it still falls behind both China and the US.
• Europe's take on AI regulation tends to be favourable, and contrary to the US and Russia, is in favour of a ban on the use of lethal autonomous weapons.

On April 11th, the European Parliamentary Research Service (EPRS) delivered a report focused on Defence and Artificial Intelligence (AI), authored by Sebastian Clapp. While it must be noted that the report does not officially represent the position of the European Parliament, it is a good source for getting an accurate idea of the main lines of debate and concern within the Parliament.

The structure of the report goes as follows: first, the author gives an overview of the current global race for AI, in the context of a renewed arms race, with the main players being the US and China, with Russia aiming to position itself as a global player as well. This section partially overlaps with our first article on AI and the arms industry, so we encourage you to check that article if you want more context. Second, the report explains the role of the EU in the global competition for AI. Third, the report assesses several ethical concerns bound to the use of automated weaponry, and also describes the different approaches that global powers have regarding the regulation of AI. Lastly, the report draws the consensus that the European Parliament seemingly holds on the use of AI in warfare.

The race for AI - the race for weapons

The use of automated weapons is not something new. Something that can be taken from this report is that global leaders became openly aware of the role of AI already in the years 2016-2018, way before AI became central in public opinion, as has happened since 2022. For instance, Vladimir Putin said in 2017 that "whoever becomes the leader in this sphere [AI] will become the ruler of the world". Among the uses of AI, we can count its relevance to logistics, intelligence gathering and analysis, or disinformation campaigns. In fact, when we think about AI and the defence industry, we tend to think about lethal autonomous weapons, also known as LAWS, but this represents only a part of all the uses of AI in this field.

The report puts four main actors in this renewed arms race: the US, China, Russia, and the EU. The document notes that the US still holds a significant edge over the rest of the countries, investing vast sums of money (US$1.8 billion only in 2024) into "more than 685 active AI projects currently underway". Notably, among the main institutions within the US we can find names such as DARPA, an agency formed back in 1958 in the midst of the cold war, but also probably less familiar names such as the Joint Artificial Intelligence Center (JAIC), formed in 2018, and now fully integrated within the Chief Digital and Artificial Intelligence Office (CDAO), which supervises from the Pentagon the US AI strategy.

In second place, unsurprisingly, we find China to be the main rival of the US. The report suggests that in the case of China, it is way more difficult to assess the amount of money the government is pouring into AI research and technology, but there are some reliable estimations. For example, a report made by the US Department of Defense released in 2024 estimates that total Chinese spending on defence might be up to an impressive amount of US$450 billion, and that a non-negligible part would be destined to AI-related projects.

And in third place we would find Russia, which, despite being way behind both China and the US, has committed to substantially modernising its armies, aiming to automate 30% of its equipment by 2025. As mentioned above, Russia clearly acknowledges the potential of AI, also beyond the military. Now, it is difficult to assess how much progress Russia has made in that regard, also because Russia's budget has not been disclosed, just as China's has not. Nevertheless, as we mentioned elsewhere, one of the particularities of the war between Russia and Ukraine is that it is the first one in which both sides are consistently using AI-powered equipment in their war efforts.

Before going to the state of the European Union regarding AI and defence, the report highlights the crutial role that some companies, such as Microsoft, Palantir, Alphabet (the holder of Google), Amazon or Apple, on the side of the US, or ByteDance (the owner of TikTok), Alibaba or Tencent, on China's side, play on the AI race. And the report rightly highlights that these companies have become key geopolitical actors. In its own words:

Today's tech giants operate with a level of autonomy unprecedented in history. They often outpace governmental agencies in innovation, particularly in AI. The influence of private tech companies extends beyond economic and technological domains into the realm of international security. Their control over critical infrastructure, data, and communication networks grants them a strategic position that can affect national security decisions.

The EU's position on the global stage

That the role of AI has reached the ears of European policymakers and companies alike is not a surprise. The EU's Strategic Compass, released in 2022, highlights the importance of investing in critical technologies, which include AI, when it comes to defence, civilian technology and, notably, space technology. Moreover, the white paper on "European Defence Readiness 2030" states that:

Geopolitical rivalries have not only led to a new arms race but have also provoked a global technology race. Technology will be the main feature of competition in the new geopolitical environment. A handful of critical and foundational technologies like AI, quantum, biotech, robotics, and hypersonic are key inputs for both long term economic growth, and military preeminence. Boosting innovation is key for this. As such, technology diffusion for commercial purposes must be reconciled with more rigid technology ecosystems to advance national security objectives. [...] The same applies to security of supply of critical raw materials, fundamental for our economic and industrial production, defence capabilities and competitiveness.

Further funding is already on its way. The European Commission's ReArm Europe plan aims to channel up to 800€ billion into its defence capabilities. Interestingly, it has been suggested that AI could also improve the efficiency of this sector, since, as the report quotes from a study carried also by the EPRS which states that "the 'cost of non-Europe' in defence spending ranges from EUR 18 billion to EUR 57 billion annually". That being said, there is no specific mention on how AI could improve governance and efficiency, and while the case can be made about potential benefits on AI in that regard, the solution, as we see it, to Europe's fragmentation and lack of projection is primarily political, not technical, since Europe has everything it needs to streamline its functioning and to become more competitive and efficient: it just requires political will. In this regard, AI can be more helpful when applied to neutralise cyber-threats and misinformation campaigns than creating better budgets. Still, the report proceeds by acknowledging that, besides the increasing awareness and investment in defence, including AI, Europe still has a long way to go. The continent has been burdened by years of underinvestment, and right now the only two European countries that are relevant in this regard are (yes, you guessed it) France, with companies like Mistral AI, and Germany, with companies like Helsing.

Now, besides increasing the investment related to defence, the EU is also aiming to foster the civilian sector. Not only because it is needed to improve Europe's competitiveness overall, but because, as the report notes, AI is primarily a civilian technology that is being transferred towards the military, unlike other inventions like the radar or the Internet, which have a military origin, and then were adopted by the civil society. In this case, a primary source of funding comes from the Horizon Europe program, but also there are other relevant iniciatives such as the EU Defence Innovation Scheme (EUDIS), which counts on EUR 1.46 billion sourced from the European Defence Fund (EDF), on the one hand, and on EUR 400-500 million leveraged from other public and private sources, on the other. There are other relevant mechanisms and institutions at play, such as the European Defense Agency (EDA), created back in 2004, and the European Defence Innovation Hub (HEDI), created to work within the EDA in order to locate and promote projects of strategic importance, as well as to provide funding for "defence innovation prizes, proof-of-concept development, European defence innovation shows, and innovation challenges".

Ethics and regulation

Finally, the report focuses on two main points: ethics and regulation. Briefly speaking, AI poses obvious ethical concerns, which range from accountability and responsibility to reliability and potential misuse of the equipment. Some cases that come to mind are, for instance, target-selection, accountability of AI-made decisions, and until which point automated warfare, by reducing a priori the number of human casualties, can create an incentive for war by reducing its cost in personnel, and therefore producing further casualties a posteriori just by increasing the number of conflicts overall.

As an example not mentioned in the report, especially concerning civilian casualties, Google has been worried by the fact that it could not control how Israel was using Project Nimbus on its deadly assault on Gaza and the occupation of the West Bank - but signed the deal anyway. For context, Project Nimbus consists of a partnership between Google and the Israeli government, according to which Google, alongside Amazon Web Services, provides cloud computing, AI and machine learning solutions to Israel; the project is currently valued at US$ 1.2 billion. If you wonder why Google appears here, it's worth mentioning that earlier this year, Google lifted its ban on the development of AI weaponry, (although these two news are not directly linked) following earlier announcements made by companies like Meta by the end of 2024.

Regarding regulation, the report contrasts the US's anti-regulation, "pro-business" position against the EU's regulatory efforts (we will issue an article on the apparent conflict between innovation and regulation, and why we generally believe that this conflict is not such). While Biden's administration, among other things, adhered to the Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy launched in The Hague during 2023, Trump's administration has taken a more anti-regulatory attitude, with US Vice President JD Vance putting pressure on the EU to lift some of its laws. Let's remember that the EU notably launched in 2023 its AI Act, the first binding worldwide regulation of its kind. As of now, the US is the only country, along with Russia, to explicitly oppose a ban on lethal autonomous weapons (LAWS), with the EU being in favour of it and China somewhat in between, supporting a ban on the use of LAWS, but not on its development.

Moving to the European Parliament, the report states that:

Overall, the Parliament is rather open to military AI use and recognises AI's strategic importance and potential to protect soldiers and civilians. Members emphasise three key principles for military AI: (1) maintaining human involvement in command and control, (2) ensuring legal accountability for individuals and states, and (3) advocating for international AI governance through the UN, including export regulations.

To this, the consensus regarding the ban on LAWS can be added.

Final remarks:

AI and its incorporation into the military are facts, and Europe seems aware of them. We believe that the EU's regulatory efforts are valuable. First, because they protect fundamental citizens' rights. Second, Europe gives large amounts of data to other countries freely, which is used to enrich companies abroad. These companies, as the report notes, are geopolitical players. Protecting the rights of European citizens is an obligation of the State to its population, but it is also in its geopolitical interest. No wonder that countries like the US adopt an anti-regulatory stance, at least in the international scene. It is not a matter of mentality, as it is often suggested, but it is a matter that US companies simply benefit from their data-related extractivist practices around the globe - and of course, it is far from the first time that something like this happens.

That being said, it is clear that the EU needs further investment and, equally important, political will to not fall behind from a technological and strategic point of view. The need for major investments is a part of Europe's quest towards its technological and digital sovereignty in a world that finds itself in a renewed arms race. In this scenario, it is crucial to ensure compliance with international law and humanitarian standards.

Precisely on this regard, the report mentions the ongoing war between Russia and Ukraine as an example of a conflict in which AI is playing an important role. But the report omits, for instance, the ongoing war between Israel and Hamas in Gaza. In this case, AI has been consistently used against both military and civilian targets, and the UN consistently condemned Israel's use of AI in the ongoing conflict.

On a different note, the report acknowledges NATO's role in Europe's development of its defensive capabilities. Still, it will be worth checking if we will see a rise of a new AI diplomacy. For instance, earlier this year, the first AI Action Summit was held in Paris and was co-chaired by French President Emmanuel Macron and Indian Prime Minister Narendra Modi. A day after the summit, the India-France Declaration on Artificial Intelligence was announced, bringing India and, in this case, France closer when it comes to AI and regulation. Moreover, France is an important arms provider to India, only behind Russia. As French companies sign contracts with different governments within the EU and also the UK, it will be worth checking if the same starts happening between EU and other non-EU countries. In any case, it is clear that India has its own path, and it could emerge as a relevant actor also in this field in the foreseeable future. In a moment when the EU is in need of new partners, new forms of diplomacy such as this can arise.