“NATO-Mation”: Strategies for Leading in the Age of Artificial Intelligence

Beyond that, we must also acknowledge that defence is not the driving force of technological developments in AI. The technology is and will continue to be made available by the commercial sector. The question is: are we agile enough to seize it? The pace of technological developments of the digital economy is of a magnitude that is totally unknown for the traditional defence acquisition systems. The real battle is therefore one of the uses, to be able to leverage volatile technologies before they become obsolete. That means that the real question for AI in the military realm is not so much the technology as it is the across-the-board transformational effort (organizational, cultural, conceptual, procedural) that it will take to reap its benefits.

We must also obviously adopt a DOTMLPFI2 capability development approach, with a specific focus on the human dimension, which is essential to operationalize AI. Man-machine teaming is key and must rely on well-designed man-machine interfaces but first and foremost on informed trust to the machine (not blind faith!) and clear decision processes – which puts the human being in the appropriate place with respect to the kind of decision and which are in compliance with our legal obligations and ethical principles. In this context, I applaud the initiative taken by Andrea Gilli at the NATO Defense College to offer such a comprehensive and thought-provoking body of work to inform the establishment of a NATO AI strategy. I am therefore honoured to preface this remarkable study.

Many of its proposals closely match Allied Command Transformation (ACT) efforts to lead the transformation of NATO’s military instrument of power, and principally ACT’s own contribution to the establishment of a NATO AI strategy. This strategy should aim at: leveraging AI to out-think, out-perform, and out-pace our potential adversaries; improving decision making at every NATO echelon; optimizing performance of priority NATO capabilities; and driving agility and continuous improvement.

Such a strategy requires efforts in various directions. We should start by sourcing commercially mature AI applications to demonstrate value and create early momentum for the adoption of AI. We must also improve our ability to pro-actively shape the AI technology investment landscape and development of the defence specificities. Moreover, NATO has to rethink its operating model and organization with a prime focus on speed.

Finally, NATO will need to scale AI with a technology and talent foundation, meaning we need to establish a NATO Infostructure.

A key condition to implement successfully these lines of effort is truly embracing innovation and agility. ACT, as one of the leading agents for innovation at NATO, has taken bold steps in this direction in the last two years. We established an Innovation Branch, a place where our innovators are “protected and nurtured”, to use the words of the “NATO-mation” study. As part of it, our Innovation Hub has been equipped with an “Open Innovation Lab” capability, implementing state-of-the-art agile methodologies (“DevSecOps”3). Our Innovation motto, “Start Small, Think Big, and Scale Up towards Full Scale Continuous Innovation”, is a concrete way to phrase the “3S” strategy approach proposed by the “NATO-mation” study.

The Lab is notably exploring currently the potential of AI to mine and analyze open source datasets to provide Intelligence communities easily-accessible, supplementary data facilitating target area risk assessment during mission planning and real time operations.

We need to build upon these successful experimentations and adopt the agile approach at the NATO Enterprise level. This is the intent of the proposals put forward with our sister command, Allied Command Operations, and of the partnership we established this summer with the NATO Communications and Information Agency (NCIA) to work towards the adoption of DevSecOps in NATO at scale.

Scaling up is also about appropriately funding innovation, an imperative that we have raised on numerous occasions and that I will continue to advocate with passion.

On another note, ACT strives to cultivate an innovative workforce through our regular “i3” (initiate, innovate, imagine) events, and bring fresh perspective to NATO through our young disruptor forum, Hackatons and Innovation Challenges.

We also explore the potential of emerging technologies through Disruptive Technology Assessment Games in order to help build concretely common understanding among the Allies, and address the interoperability issues upstream via our TIDE Sprint events. We have established a pool of expertise on data within ACT, notably missioned to advocate for NATO to become a Data Driven Organisation.

Last but not least, we support NATO HQ to progress on the key ethical questions raised by the adoption of AI for military purposes. On this topic, I am convinced that we will identify these issues as we progress – we are just at the beginning of our journey. My message to our political leaders is that the ethical reflections are essential but must not slow us down in the exploration of the potential of this technology and in our R&D efforts.