You don’t need to be an expert in frontier AI or AI safety, but having some knowledge and expertise on AI could mean you'd be of great help to the UK Civil Service.
The technical talent gap in UK policy
Most of the policymakers working on AI regulation are not subject-matter experts, and many are new to tech policy. This is not meant as a criticism, but as a fact. Civil servants are generalists and as such are hired for general competency, and not because they’re subject matter experts. However, AI is quite a deeply technical subject in many regards, and therefore subject-specific input could greatly help the quality and outcomes of policy decisions.
Subject-specific input to government policy typically happens by:
Policymakers having access and engaging with to advisers, which could look like external field experts, researchers, and lobby groups.
However, with the recent creation of the AI Safety institute and scaling up of AI policy in departments including the 'Department for Science, Innovation & Technology', 'Cabinet Office', and various others - there is a fantastic opportunity for field experts to join the civil service directly.
There are good reasons why the UK Civil Service may be a very strong option for people with AI safety knowledge
The UK is rapidly moving on policy.
There seems to be a real understanding and embracing of the full pace of AI development, and there are many bastions of hardworking, vocational policymakers in the UK civil service.
You are likely to have responsibility for important and interesting work, even at an early stage.
Many areas of the civil service provide great opportunities for having considerable influence and responsibility. Speak to any civil servant, and especially those that have experience in DSIT's recent AI teams, and you'll hear tales of regularly being asked to draft important documents about how the AI should regulate policy to key decision makers. This means your work could very quickly be landing on the desk of the Secretary of State, other Cabinet Ministers, or inside the No.10 itself.
The UK will have a Brussels effect on US policy.
There is no escaping that the US matters a lot, especially on issues related to tech. However, there is a good argument that a window of time exists where the UK is ahead on AI policy, and there could be a mechanism for a Brussels effect via things like the potential collaboration between the UK AIS Institute and the US one.
Some reasons it’s not a strong option for everyone
There will be some bureaucracy.
Large organisations are often bureaucratic, but developing excellent AI policy involves multiple bureaucracy. There’s no magic future where any nation does away with bureaucracy in political institutions (having input from multiple stakeholders is after all part of the democratic process!).
There will be a lot of working with people.
There's often lots of need to engage stakeholders and have meetings to discuss policy options and different problems.
You may feel uncertain about the impact of your actions.
Unfortunately, the policymaking process is opaque, even to the people involved in it. There’s no magic answer to “the most impactful role to have in AI governance”. Joining the civil service is one way to become a part of the conversation and the journey.
The way to make a successful application to the civil service is to ensure you demonstrate the behaviours it assesses for with your past experiences during the application and interview.
IGC provides bespoke support to help you through the application process, with feedback on your behaviours and interview techniques. Reach out to us now to request for mentoring support.
NB. External collaborators helped draft this blog.