Navigating Ethical Challenges in AI and Defense

Arushi Agarwal from the European Sustainability Strategy team and Aerospace & Defense Analyst Ross Law unpack what a reshaped defense industry means for sustainability, ethics and long-term investment strategy.

Read more insights from Morgan Stanley.

----- Transcript -----

Ross Law: Welcome to Thoughts on the Market. I'm Ross Law from Morgan Stanley's European Aerospace and Defense team.

Arushi Agarwal: And I'm Arushi Agarwal from the European Sustainability Research Team.

Ross Law: Today, a topic that's rapidly defining the boundaries of sustainable investing and technological leadership – the use of AI in defense.

It's Tuesday, July 8th at 3pm in London.

At the recent NATO summit, member countries decided to boost their core defense spending target from 2 percent to 3.5 percent of GDP. This big jump is sure to spark a wave of innovation in defense, particularly in AI and military technology. It's clear that Europe is focusing on rearmament with AI playing a major role.

In fact, AI is revolutionizing everything from unmanned systems and cyber defense to simulation training and precision targeting. It’s changing the game for how nations prepare for – and engage in – conflict.

And with all these changes come serious challenges. Investors, policy makers and technologists are facing some tough questions that sit at the intersection of two of Morgan Stanley's four key themes: The Multipolar World and Tech Diffusion.

So, Arushi, to set the stage, how is the concept of sustainability evolving to include national security and defense, particularly in Europe?

Arushi Agarwal: You know, Ross, it's fascinating to see how much this space has evolved over the past year. Geopolitical tensions have really pushed national security much higher on the sustainability agenda. We're seeing a structural shift in sentiment towards defense investments.

While historically defense companies were largely excluded by sustainability funds, we're now seeing asset managers revisiting these exclusions, especially around conventional and nuclear weapons. Some are even launching thematic funds, specifically focused on security and resilience.

However, in the absence of standard methodologies to assess weapon related exposures, evaluate sector-specific ESG risks and determine transparency, there is no clear consensus on what sustainability focused managers can hold.

Greater policy focus has created the need to identify a long-term approach to investing in this sector, one that is cognizant of ethical issues. Investors are now increasingly asking whether rapid technological integration might allow for a more forward-looking, risk aware approach to investing in national security.

Ross Law: So, it's no news that Europe has historically underspent on defense. Now, the spending goal is moving to 3.5 percent of GDP to try and catch up. Our estimates suggest this could mean an additional $200 billion per year in additional spend – with a focus on equipment over personnel, at least for the time being. With this new focus, how is AI shaping the European rearmament strategy?

Arushi Agarwal: Well, AI appears to be at the core of EU’s 800 billion euro rearmament plan. The commission has been quite clear that escalating tensions have not only led to a new arms race but also provoked a global technological race.

Now to think about it, AI, quantum, biotech, robotics, and hypersonic are key inputs not only for long-term economic growth, but also for military pre-eminence.

In our base case, we estimate that total NATO military spend into AI applications will potentially more than double to $112 billion by 2030. This is at a 4 percent AI investment allocation rate. If this allocation rate increases to 10 percent as anticipated by European deep tech firms, then NATOs AI military spend could grow sixfold to $306 billion by 2030 in our bull case.

So, Ross, you were at the Paris Air Show recently where companies demonstrated their latest product capabilities. Which AI applications are leading the way in defense right now?

Ross Law: Yeah, it was really quite eye-opening. We've identified nine key AI applications, reshaping defense, and our Application Readiness Radar shows that Cybersecurity followed by Unmanned Systems exhibit the highest level of preparedness from a public and private investment perspective.

Cybersecurity is a major priority due to increased proliferation of cyber attacks and disinformation campaigns, and this technology can be used for both defensive and offensive measures. Unmanned systems are also really taking off, no pun intended, mainly driven by the rise in drone warfare that's reshaping the battlefield in Ukraine.

At the Paris Airshow, we saw demonstrations of “Wingman” crewed and uncrewed aircraft. There have also been several public and private partnerships in this area within our coverage. Another area gaining traction is simulation and war gaming. As defense spending increases and potentially leads to more military personnel, we see this theme in high demand in the coming years.

Arushi Agarwal: And how are European Aerospace and Defense companies positioning themselves in terms of AI readiness?

Ross Law: Well, they're really making significant advancements. We've assessed AI technology readiness for our A&D companies across six different verticals: the number of applications; dual-use capabilities; AI pricing power; responsible AI policy; and partnerships on both external and internal product categories.

What's really interesting is that European A&D companies have higher pricing power relative to the U.S. counterparts, and a higher percentage are both enablers and adopters of AI. To accelerate AI integration, these companies are increasingly partnering with government research arms, leading software firms, as well as peers and private players.

Arushi Agarwal: And some of these same technologies can also be used for civilian purposes. Could you share some examples with us?

Ross Law: The dual use potential is really significant. Various companies in our coverage are using their AI capabilities for civilian applications across multiple domains. For example, geospatial capabilities can also be used for wildfire management and tracking deforestation. Machine learning can be used for maritime shipping and port surveillance.

But switching gears slightly, if we talk about the regulatory developments that are emerging in Europe to address defense modernization, what does this mean, Arushi, for society, the industry and investors?

Arushi Agarwal: There's quite a lot happening on the regulatory front. The European Commission is working on a defense omnibus simplification proposal aimed at speeding up defense investments in the EU. It's planning to publish a guidance notice on how defense investment will fit within the sustainable finance framework. It’s also making changes to its sustainability reporting directive. If warranted, the commission will make additional adjustments to reflect the needs of the defense industry in its sustainability reporting obligations.

The Sustainable Fund Reform is another important development. While the sustainability fund regulation doesn't prohibit investment into the defense sector, the commission is seeking to provide clarification on how defense investment goals sit within a sustainability framework.

Additionally at the European Security Summit in June, the European Defense Commissioner indicated that a roadmap focusing on the modernization of European defense will be published in autumn. This will have a special focus on AI and quantum technologies. For investors, whilst exclusions easing has started to take place, pickup in individual positioning has been slow. As investors ramp up on the sector, we believe these regulatory developments can serve as catalysts, providing clear demand and trend signals for the sector.

Ross Law: So finally, in this context, how can companies and investors navigate these ethical considerations responsibly?

Arushi Agarwal: So, in the note we highlight that AI risk management requires the ability to tackle two types of challenges. First, technical challenges, which can be mitigated by embedding boundaries and success criteria directly into the design of the AI model. For example, training AI systems to refuse harmful requests.

Second challenges are more open-ended and ambiguous set of challenges that relate to coordinating non-proliferation among countries and preventing misuse by bad actors. This set of challenges requires continuous interstate dialogue and cooperation rather than purely technical fixes.

From an investor perspective, closer corporate engagement will be key to navigating these debates. Ensuring firms have clear documentation of their algorithms and decision-making processes, human in the loop systems, transparency around data sets used to train the AI models are some of the engagement points we mention in our note.

Ultimately, I think the key is balance. On the one hand, we have to recognize the legitimate security needs that defense technologies address. And on the other hand, there's the need to ensure appropriate safeguards and oversight.

Ross Law: Arushi, thanks for taking the time to talk.

Arushi Agarwal: It was great speaking with you, Ross,

Ross Law: And thank you all for listening. If you enjoy Thoughts on the Market, please leave us a review wherever you listen and share the podcast with a friend or colleague today.