Artificial Intelligence for Our Defence and Security

The British Ministry of Defense (MoD) has today launched the Defense Artificial Intelligence Strategy at the AI Summit London. In this article, MoD second permanent secretary Laurence Lee explains its significance for defense and beyond.

Today’s AI Summit marks a truly exciting day for defense as we launch the defense Artificial Intelligence Strategy.

AI is one of the greatest technological revolutions of our lifetime, with huge potential for global prosperity, security, and social advancement. As the 2021 NATO AI strategy stated: “AI is changing the global defense and security environment. It offers an unprecedented opportunity to strengthen our technological edge but will also escalate the speed of the threats we face”.

The strategy reflects the wider backdrop of escalating international tensions. We have seen the greatest resurgence of threats to our national security since the end of the Cold War. We are presented with an uncertain threat environment, ranging from hostile states – as the Russian invasion of Ukraine makes clear – and non-state threats, such as terrorist groups and transnational organized crime. Adversaries will increasingly harness AI to enhance cyber and information warfare and accelerate military operations to machine speed.

Our response in the AI Strategy is ambitious and comprehensive, calling for a once-in-a-generation technological shift across every part of our business. Our vision is that in terms of AI, U.K. defense will be the world’s most effective, efficient, trusted, and influential defense organization for our size. We intend to exploit AI fully to revolutionize all aspects of defense business, from enhanced precision-guided munitions and multi-domain Command and Control to machine speed intelligence analysis, logistics and resource management.

U.K. expertise in AI

We know we can’t do this alone and recognize that the U.K. AI sector is a strategic national resource – as vital now as coal and steel were in the industrial age. AI expertise and innovation is found across a broad ecosystem – from some of the largest companies in the world to university spinouts and fast-moving small-to-medium scale tech enterprises.

Private tech sector investment and innovation far exceeds that of governments. For defense to retain our technological edge over potential adversaries, we must partner with industry and increase the pace at which AI solutions can be adopted and deployed throughout defense. There is a huge opportunity for two of our leading industries, the defense and security sector and the AI industry, to jointly create world-leading AI capability, furthering our prosperity, and protecting our security.

To make these partnerships a reality, defense will establish a new Defense and National Security AI Network, clearly communicating our requirements, intent, and expectations and enabling engagement at all levels. We will establish an industry engagement team in the Defense AI Centre to enable better defense understanding and response to the AI sector. It will also promote the best and brightest talent and exchange of expertise between defense and industry.

Furthermore, defense will work to reduce the barriers that prevent viable companies – particularly SMEs – from entering our supply chains. Access to defense datasets and facilities such as our specialist computing capabilities and test and assurance infrastructure will be simplified. Defense will also work with cross-government partners, suppliers, and international partners to protect critical AI capabilities and Intellectual property.

The need for trust

However, these efforts will fail if the AI sector does not want to work with us. We understand that AI is not infallible, that a range of challenges and issues – from algorithmic bias to the development of unpredictable behavior over time – can lead to undesirable or unethical outcomes. We also understand that AI-related risks and concerns can be particularly acute in a defense context, and that academics and private sector partners rightly demand confidence that they are working with a responsible actor.

We know this must be a partnership based on trust. Although we need capability which operates as intended and reflects the values of the society we serve, we must nevertheless earn that trust – and we are determined to lead by example, starting with the publication of details of the approach we will adopt to ensure our use of AI is safe, legal, ethical and responsible.

We will set clear organizational intent and ambition for the adoption and exploitation of AI, backed up with defined roles and responsibilities, as set out in the Defense AI Strategy; we will continue to apply our robust safety and regulation regimes; we will always comply with our national and international legal obligations; and we will set out a clear framework and processes for ensuring ethical adoption of the technology.

Ethics

The importance we attach to AI ethics can be seen in the extensive process we undertook to develop AI ethical principles for defense, published today. These were developed through a year-long process of consultation and challenge in partnership with the Government’s Centre for Data Ethics & Innovation, supported by our AI Ethics Advisory Panel which comprises defense officials and – crucially – external experts.

Our five principles – human centricity, responsibility, understanding, bias and harm mitigation, and reliability – are being embedded throughout all stages of any defense AI system’s life cycle. Throughout this process and into the future, we intend that engagement and transparency will be at the heart of our approach. To that end, we will publish further details of the methodology underpinning the principles in the autumn.

Deepening partnerships

The realization of AI’s contribution to protecting the U.K., our people and our values will require diligent ongoing partnership. The U.K. AI sector has immense technical strengths, human capital and is recognized internationally as an AI powerhouse, including in fields linked to the development of trustworthy AI. We are keen to access those strengths and aware that we have much to learn from each other.

By deepening the partnership between the world-leading British defense and AI industries, as set out in the AI Strategy, we can and will promote increased prosperity whilst ensuring our mutual security. I look forward to engaging with you now and in the future. Do feel free to contact any MoD colleague at this summit or during the VIP drop-in session.

Laurence Lee

MoD second permanent secretary