In national security, governments still struggle to work with startups

AUKUS governments began 25 years ago trying to draw in a greater range of possible defence suppliers beyond the traditional big contractors. It is an important objective, and some progress has been made, but governments still lack key processes for getting technologies from disparate new suppliers into service.

This is leaving security gaps. Even if they are building economical dual-use products that should be easily adapted for national security purposes, many companies find it hard to work with the government.

For example, an intelligence service may be better able to detect a hostile state threat if it could use an investigative tool already sold to banks for monitoring financial transactions. But if working for the government takes too long, the vendor will simply stick with its financial customers.

The issues are complex, but the ball is in governments’ court. If governments want to benefit from vendor diversification in defence and security, they’ll have to be more open on the problems they are trying to solve, quicker to contract with startups and more willing to learn through rapid iteration.

Governments can seek to solve their hardest security-technology challenges in a few ways. They have traditionally tried to build the solution in house or have paid a large defence prime to do it. This century they’ve slowly realised that their hard problems are not always unique and that solutions already exist in the marketplace—just not within their traditional supplier base.

In 1999, the CIA set up In-Q-Tel and in 2004 US Grand Challenges were established; these were competitions open to many entrants, often newcomers. Britain has published technology challenges since 2008 and recently set up a Co-Creation Centre. Australia joined In-Q-Tel International in 2018 and has similar programs to engage the startup community.

That’s a lot of effort in the right direction. Yet governments still find it hard to engage with non-traditional suppliers. One reason stands out: officials trying to pull-through new capabilities have to fight against the system to take a risk on a novel supplier.

Venture capitalists, on the other hand, are prepared to take risks early, backing a company with millions of dollars based on potential sales. Many Silicon Valley investors actually discourage startups from trying to sell to governments, even if the products are directly relevant to national security or defence. Government sales are too sporadic and slow for them.

Unfortunately, the best choice for many companies is often to stick with making products for commercial customers and just being ready for government customers to turn up if they want to. The attitude is: ‘build it and (maybe) they will come.’

Governments would ideally be able to exploit a product for defence and security operational advantage before it reaches the open market. They now regretfully acknowledge that game-changing products are being sold before they can be exploited in secret.

The best approach to solving this is countercultural to organisations who want to keep everything under wraps secret and built in-house. The answer is to be more open, less secretive and to accept more risk.

Governments openly talk about what technologies they care about in the longer term (such as quantum or synthetic biology) and how they are improving what’s called ‘pull-through’—the gap between technology demonstration and deployment. The mechanisms include grand challenges, grants, co-investments and quick contracting.

But governments are not good at explaining why some technologies are important and where they will be deployed operationally, often because of a preference for secrecy. If a government takes years to build a technology platform in secret, the result is likely to be secure but will be quickly overtaken by commercial technology.

The conflict in Ukraine means military technology is being conceived, developed, tested and deployed at eye-watering speeds in front of social media and the public. Governments are only now being more open about the operational context in which the technology will be deployed, share specific requirements and constraints openly and upfront (that is, explain the why); meaning the right solution can be developed more quickly.

Britain is leading the way. For instance, the Ministry of Defence has publicly tendered for one way attack drones and autonomous maritime weapons for Ukraine.

Sharing of why and where technology will be deployed results in quicker operationalisation. New and better versions can be created with user feedback; the first version can be discarded and replaced with a new one, which itself might not last long as iterations appear. In this paradigm, the rapidity of pull-through becomes a source of operational security, because the adversary is forever trying to catch up on what you’re doing.

With more global competition and the West’s diminishing technological superiority demands government embrace proper processes to harness dual-use commercial technology—of the speed and scale that nimble companies normally focused on private customers can deliver. We must make it work.