War in the age of AI demands new weaponry

Stay informed with free updates

Simply sign up to the Aerospace & Defence myFT Digest — delivered directly to your inbox.

The writer is former CEO of Google, chair of the Special Competitive Studies Project, an honorary KBE and founding partner of Innovation Endeavors

The defence industry is having a moment. As the world becomes more dangerous, with Russia’s war in Ukraine, conflicts in Gaza and beyond, and the backdrop of great power competition, defence budgets across the globe are rising.

Exact figures are hard to discern, but estimates show global military expenditure has increased 34 per cent over the past five years. The five US prime defence contractors have all received significant new orders.

This defence boom coincides with another technological revolution unfolding all around us: artificial intelligence. As increased defence budgets meet the AI revolution, procurement decision makers should favour weapons systems that are affordable, attritable and abundant. As a result, many more opportunities are coming for start-ups and defence unicorns.

I’m investing in such start-ups because we need new capabilities to compete in this changing world. While there are substantive differences between the US and Ukrainian military, there are a number of important lessons to be learnt from the country’s conflict with Russia. Here are three that should help guide the military, our defence appropriators and our procurement specialists as we navigate the future. 

First, “you go to war with the army you have, not the one you want”. That’s why it’s so important that the increase in defence spending and replacement of arms sent to Ukraine do not simply reload US stockpiles but retools them and the defence industry that supplies them. While the margins on expensive “exquisite” systems and maintenance will enable prime defence contractors to buy back stock, more money should go into research and development. This could enable us to one day supplement, if not replace, F-35 fighters and the apparatus around them — including mid-air refuelling aircraft and ground staff — with long-range autonomous drone units. At a minimum, the profits could go to acquire new companies, bolstering the incentive structure for defence start-ups building a different future of American weaponry.

A second lesson is that we need systems that can communicate effectively even in challenging environments, particularly those saturated with electronic warfare. A clear lesson from the Ukrainian front has been the challenge of operating amid GPS denial and spoofing, while targeting enemy forces 100km or more away. In such scenarios, traditional electronic navigation systems can become compromised, leaving missions incomplete and forces vulnerable. We need weapons and systems that can function reliably even when conventional methods fail, including GPS alternatives like quantum navigation and visual odometry. These technologies are not just about surviving in a contested environment but ensuring we have an information advantage over our adversaries in the most critical moments. We need to ready ourselves to procure what we need to fight whenever and wherever, which in this new age will most likely be overseas in places filled with electronic warfare. 

The final lesson is the impact of asymmetric warfare in creating disparities in the cost-capability ratio, which is only accentuated the longer a conflict persists. As we’ve seen in Ukraine, it is unsustainable to fire a $4mn patriot missile to intercept a $50,000 Shahed drone. The canonical example in the US is the fear that a few $10mn-$20mn Chinese DF-21 or DF-26 anti-ship ballistic missiles could destroy a US aircraft carrier, its equipment and its 6,000 member crew — all for less than 1 per cent of the carrier’s $13bn price tag. To combat these fears, we need cheaper, more numerous alternatives that take advantage of interconnected and nimble software. But this will require procurement reform that moves beyond best price contracts and path dependency. Rather, we need more sophisticated purchasing approaches that weigh total cost and supply chain resilience with other factors like performance and adaptability. 

There are certain headwinds that could keep defence budgets from rising indefinitely. In the US, worryingly, debt repayment costs surpassed annual defence spending for the first time in 2024. Even within defence spending, personnel costs have grown considerably, risking the crowding out of opportunities for modernisation and new weapons development.

Still, the odds are the US will continue to grow its spending. Senator Roger Wicker, the ranking member on the Senate armed services committee, has proposed a pathway to return defence spending to nearly 5 per cent of GDP, levels not seen since the 2009 surges in Afghanistan and Iraq. That would hasten the arrival of the first-ever trillion-dollar US defence budget.

While the majority of this money would likely focus on programmes of record and sustaining the rising cost of healthcare and pension costs for our soldiers, sailors, air force and guardians, there are some glimmers that the US is taking innovation seriously. The country’s National Defense Authorization Act, which authorises funding levels for the US military, envisioned a fivefold increase in the budget for the Defense Innovation Unit in the last year to almost $1bn. While that’s far less than half of 1 per cent of the total, it is a start. 

And it’s a start that is echoing around the world. In recent years, the number of Nato countries meeting or surpassing their 2 per cent of GDP target for defence spending rose from nine in 2020 to 23 this year. New multinational initiatives are emerging too, such as Nato’s Diana and Innovation Fund, which funds AI technologies across the west.

We must think strategically about the objectives we target, and the best way to achieve them — not simply how we achieved them in the past with existing weapons systems. We may have a lot of Abrams tanks but not every problem is a nail. This is why innovation is so important. We cannot waste this opportunity simply buying the same weapons we fought with in our past wars. The age of AI demands that we invent, adapt and adopt the weapons of AI.