‘Do the Right Thing’ – The Dilemma of Tech Companies and Research Institutes Involved in Defence Innovation

Flora Matthaes discusses the dilemma that research and development institutions face when they are confronted with tenders from the defence sector.

The contribution of tech companies and research institutes to the development of new military technologies is in the spotlight again. In March, media outlets uncovered that tech giant Google provided Artificial Intelligence (AI) expertise for a Pentagon defence project seeking to improve image processing by Unmanned Aerial Vehicles (UAV), commonly known as drones. The ‘Maven Project’ is developing technologies to support the US ‘Defeat-ISIS Campaign’ in Iraq and Syria.

This revelation sparked a controversy. Some Google employees allegedly resigned over moral concerns. Many more signed an open letter protesting the firm’s “business of war”. The company’s current slogan ‘Do the right thing’ was cited as evidence for an apparent contradiction between its pacifist corporate culture and involvement in a defence project. As a result, Google has reportedly opted against renewing the Pentagon contract. Nonetheless, the debate about the appropriateness of linking commercial and security interests remains.

From California to Brussels

In technology research and development (R&D) the above case is not unique. Firms like Amazon, Microsoft, IBM and Oracle have been mentioned in relation to the tender process for a multi-billion cloud infrastructure catering to US military needs on and off the battlefield. In August 2017, then US Defense Secretary Mattis visited the ‘Defense Innovation Unit-Experimental’ (DIUx) in Silicon Valley, where the US government facilitates investments into commercial start-ups with disruptive defence technological potential. To channel AI efforts including ‘Maven’, the US Department of Defense (DoD) created the ‘Joint Artificial Intelligence Center’ (JAIC) last month.

Across the Atlantic, the EU has recently awarded its first-ever grants to exclusively defence R&D projects, to be followed by a more long-term and larger funding programme. Similar to the US example, this step has also attracted criticism. Scientist activists warned against the risk of a growing “militarisation” of the EU, which received the Nobel Peace Prize in 2012. They furthermore expressed concerns over a possible contribution to an arms race in yet little regulated robotic weapons. A group of German universities pointed to a civil clause in their statutes, which discourages research for military purposes. This voluntary civil commitment would largely preclude applications for the new EU defence R&D funds.

Mapping the Dilemma

Although the connection between civil and military innovation is undeniable, its nature has changed over the course of time. Formerly, publicly funded research from the military realm provided innovative ‘spin-ins’ to the civil sector. Success stories include the Global Positioning System (GPS) and early computer networks. Nowadays it is mainly ‘spin-offs’ from the private sector, especially in Information and Communication Technology (ICT), which find their way into the military domain. Among these technology transfers are developments in the areas of deep learning, supercomputing and data analytics.

Accordingly, most governments insist on the absolute necessity of leveraging civil-military synergies to maintain the competitiveness of their armed forces. This is because the technological lead today is often held by the private sector, which is difficult to rival in terms of its high salaries and flexible work structures. Another argument is that other defence powers like China or Israel are already investing significant amounts of money into AI and its military application. Opting out of the ongoing technological competition would mean losing a say over its future direction.

The legal framework for this technological push, however, is underdeveloped. Especially contested is the R&D of autonomous weapon systems (AWS). According to the International Committee of the Red Cross (ICRC) definition, AWS “can select and attack targets without human intervention”. While proponents indicate the capacity of machines to reduce human error and casualties in combat, opponents fear a lack of ethical oversight in potential life-or-death decisions. Projects like ‘Maven’, which uses AI to detect objects in video imagery, could be the predecessors to such a scenario. International efforts have been made to regulate the lethal deployment of AWS (see for instance the UN Certain Conventional Weapons Convention), but they are still in their infancy.

Consequently, a dilemma emerges. Tech companies and research institutes must decide between accepting lucrative military contracts and serving national security interests or preserving their strictly civilian purpose and reputation through refraining from engaging in the R&D of next-generation weapons.

Towards a More Informed Debate

Where tech companies and research institutes decide to focus their activities is ultimately up to management. When broader technological developments with profound implications for future society arise, however, these decisions may well become a matter of public concern. The recent backlash against tech companies indicates a powerful scepticism towards the combination of commercial and security interests, which cannot easily be discarded.

Indeed, following the ‘Maven’ controversy, Google announced a new set of ethical principles to guide its development and application of AI. If companies desire to fully address existing public concerns, however, they need to move beyond proposing corporate guidelines. As far as national security considerations allow, tech companies and research institutes should publish the actual extent of their financial and technological involvement in defence innovation projects.

Increased transparency would mean those working for tech companies and research institutes could make more informed decisions regarding their employer and to which purpose intellectual property is put to use. Finally, more publicly available information on corporate activities could help rather than harm the case for responsible civil-military technology transfers by moving the current heated debate to more factual grounds.

Flora Matthaes is a graduate of International Relations at Leiden University.

Photo by S.C. Air National Guard