How to Train Diplomats to Deal With AI and Data? (WebDebate #58)

Event summary

Framing the question

Starting the conversation, Ambassador Amr Aljowaily issued a reminder that we are not dealing with a completely new question. The impact of digital information and communication technology (ICT) on diplomacy has been felt for many years and in particular over the last 30 years with the internet starting to play a more prominent role. His own publication International Relations in the Information Age (1996) is a good example and so is Jovan Kurbalija’s Information Technology and Diplomacy in a Changing Environment (1996). 

A useful typology to start thinking about capacity-development needs is to look at the various relations of AI and diplomacy, i e. AI as a topic of diplomacy, AI as a tool for diplomacy, and AI as a factor that changes the environment in which diplomacy is practised. 

In addition, it is useful to further unpack both AI and diplomacy. AI can be further dissected into questions around data, algorithms, and computational power. It is also useful to focus on distinct diplomatic functions: information gathering, negotiation, representation, and communication.

In addition, different training priorities can be identified: 

• Training for utilising AI in what it can do right, and thereby maximising its benefits
• Training for cross-checking AI in areas where things can go wrong (e.g. misinformation)
• Training for what AI cannot do at all by focusing on human capacities and skills

Specific diplomatic tasks and associated capacity-development needs

Prateek Sibal stressed that diplomats are working in global public goods. Global public goods are non-rivalrous and non-excludable. They are available to all countries and all people, and have an international dimension. Digital global public goods are of particular concern. 

In this regard, three particular tasks of diplomats can be identified. First, diplomats are setting international standards and agreements in order to reduce friction in trade and commerce (e.g. the free flow of data). Second, diplomats are mitigating global risks (e.g. cybercrime). Third, diplomats work to build and pool resources together (e.g. open source and open data). 

To do this effectively, particular competencies, which can be represented in a seven-part competency framework, can be identified (compare also Artificial Intelligence and Digital Transformation: Competencies for Civil Servants). 

In general terms, diplomats need to develop:

1. A capacity for systems thinking
2. Strategic foresight
3. Agility (in designing responsive international agreements, for example)

Specifically in the area of AI, diplomats need: 

4. Data literacy
5. A good understanding of issues around privacy and security
6. Knowledge of interoperability and its legal, ethical, and regulatory principles
7. The ability to apply people-centric approaches to technology. 

Designing effective training and capacity development

Claudia del Pozo highlighted a number of examples that focused on in-vivo and experiential learning. Meta’s Open Loop programme, led in collaboration with C Minds' Eon Resilience Lab in Mexico and Uruguay, is an example that brings technologists and policymakers together to explore regulation options in a conversational and explorative setting to improve future regulation.

Similar programmes could be imagined with diplomats. A key lesson from such examples is the effectiveness of project-based and peer learning, where knowledge can be freely exchanged across multiple stakeholder groups. 

In addition, general training on AI should be implemented into formal diplomatic training, for example at diplomatic training academies and similar institutions.

Broader shifts and concerns

Aljowaily highlighted the need of being mindful of the broader shift in the work of diplomats. For example, within the communication function, he identified a shift from being a representative to being an executive. Another shift from being ‘only’ a representative towards being more of an executive could be identified in the representation function. 

Del Pozo stressed the importance of AI ethics as a topic that should be included in all aspects of training and capacity development in this area. 

Speakers agreed on the importance of inclusiveness (including gender, women, and youth), and stressed the importance of assisting developing countries to be heard in various global conversations, bridging the digital divide, and addressing divides within societies.

Speakers wondered whether diplomats need to be knowledgeable on all the topics suggested. Del Pozo pointed out that it is worth shifting the focus from the individual towards a focus on teams with relevant skill sets. 

Sibal issued a reminder of the AI alignment problem: how can we align technology with human values and priorities. 

In closing, Aljowaily highlighted a number of continuous challenges: 

• How to ensure that developing countries benefit from new digital ICT?
• How to re-organise ministries of foreign affairs appropriately?
• How to adopt scientific approaches into foreign policymaking?
• Where are the chances to leapfrog stages of development?