30 September 2025
CO25199 | The Role of AI in Modern Diplomacy
SYNOPSIS
Artificial Intelligence (AI) is set to become a game-changer in the field of diplomacy. From crisis response to conflict prevention and resolution, AI holds significant potential to enhance core diplomatic functions. However, its limitations necessitate a cautious, human-centred approach. For small states like Singapore, AI presents a valuable tool to enhance strategic foresight and resilience amid growing geopolitical volatility.
COMMENTARY
On his maiden visit to Singapore as the UK’s Foreign Secretary, David Lammy outlined his government’s vision for diplomacy in the digital age, stating that Artificial Intelligence (AI) will “transform how diplomacy is conducted”, because, by 2025, “diplomacy needs machine speed”. He outlined a future in which AI will revolutionise how nations project influence, anticipate threats, and forge alliances in an increasingly volatile world. He asserted that AI would transform how diplomacy is conducted, which will always have a human touch at its heart.
For millennia, diplomacy has been a fundamentally human endeavour, crafted through dialogue, negotiation, and the navigation of competing interests. Today, AI is ushering diplomacy into a new era.
AI enables rapid data processing, predictive analytics, and real-time decision support. These present opportunities for enhancing a variety of diplomatic functions, including crisis response, conflict prevention, and conflict resolution.
However, its limitations in diplomacy remain apparent, especially in its algorithmic biases and its inability to replicate the role of human intuition in tense settings. A measured approach is needed to integrate AI capabilities into the diplomatic domain.
Data-Driven Diplomacy
AI’s most significant comparative advantage is its ability to process vast datasets at speeds far beyond human capabilities. This enables the automation of time-consuming administrative tasks, such as document drafting and schedule management, freeing up diplomats to focus on strategic and relational functions.
As former US Secretary of State Antony Blinken noted, AI tools deployed in the State Department have significantly reduced the burden of manual tasks, such as summarising and translating diplomatic cables.
AI-powered predictive models are also used to predict conflicts or instability among countries. Researchers at Munster Technological University developed a model combining Bayesian Deep Learning and Random Forest techniques to forecast the stability of any country by analysing its historical and qualitative data, including traditional and social media. Predictions from diverse data inputs are generated, serving as early warning mechanisms for policymakers to engage in pre-emptive diplomacy or mediation.
Negotiation and Communication
AI models can also transform the architecture of negotiations, especially in designing peace agreements and engineering communication between parties. Trained on data from strategy games, media analyses, and historical peace agreements, the model can generate draft proposals for conflict resolution. Users input preferences across four areas: territory, security, justice, and economic arrangements. AI then evaluates the negotiability of each component for major stakeholders in the conflict.
In Libya’s peace process, the UN partnered with the AI platform Remesh to host large-scale, anonymised digital consultations with up to 1,000 participants per session. Participants could respond to key questions on governance, security, and elections, with the platform identifying areas of consensus and disagreement in real-time.
Accessible via low-bandwidth mobile devices and supporting Arabic dialects, the tool helped to shape the post-conflict roadmap, demonstrating how AI can enhance inclusion and ground peace talks in public input.
Language barriers have long complicated international negotiations, forcing diplomats to rely on interpreters who must navigate nuance, tone, and cultural subtext. AI-powered translation tools, such as DeepL, which offer real-time speech-to-text conversion, can facilitate smoother multilateral discussions.
These platforms also incorporate natural language processing, which can detect rhetorical shifts, tone, or implicit cues in speech. This can provide negotiators with deeper insights into their counterparts’ positions, particularly valuable in high-stakes settings where miscommunication can risk escalating conflict.
Crisis Management
AI can assist crisis management by analysing satellite imagery, social media trends, and news reports to predict displacement patterns and optimise manpower and resource distribution. The World Bank, for instance, has developed an AI-driven model to predict refugee influxes from South Sudan and the Democratic Republic of Congo into Uganda, drawing on hard data and online information related to the countries. Such a model can enable public services to be scaled up months before the refugees’ arrival, while also allowing for refugee integration into national services and local planning processes.
AI can also power intelligent Decision Support Systems (DSS). These systems aim to simulate optimal responses and propose best-practice solutions during high-stakes scenarios. However, DSS technologies are still in their early stages and present significant limitations. For example, the Futures Lab found that LLaMA, a generative AI model, opted for the use of force in 45 per cent of simulations. This highlights the need for further training and calibration to ensure such systems promote conflict de-escalation. Ultimately, DSS must be able to translate natural language inputs into reliable outputs that align more closely with human intent and ethical boundaries.
Simulation and Scenario Testing
Technologies like Extended Reality (XR) and digital twins, virtual models of real-world environments, allow diplomats to simulate complex negotiations. These tools test policy outcomes in risk-free settings, identifying unintended consequences before commitments are made. For example, a digital twin of a maritime dispute can simulate the outcomes of actions such as naval blockades or no-fly zones, projecting their economic and political ripple effects.
One such initiative, North Star, developed by tech start-up Anadyr Horizon, assigns probabilities of conflict or escalation to a wide range of decisions, helping decision-makers visualise risk in real time. These simulations not only enhance crisis preparedness but also support long-term scenario planning, enabling states to rehearse responses to hypothetical crises and anticipate the secondary effects of major geopolitical developments.
Limitations and Concerns
Despite the potential of AI systems in diplomacy, Professor Stuart Russell of the University of California, Berkeley, cautions against ”AI solutionism”, the view that AI is a universal problem solver. Geopolitical issues often involve deeply complex and sensitive human conflicts, where diplomacy still relies on trust and empathy, qualities that are uniquely produced through human-led interaction.
The 1970s US-China rapprochement, for instance, relied heavily on the personal rapport between Henry Kissinger and Zhou Enlai, and no algorithm could replicate the subtlety of their backchannel negotiations. While AI can provide data-driven insights, the final judgment must remain with diplomats attuned to historical context and emotional undercurrents.
Additionally, AI tools are also vulnerable to adversarial manipulation. Deepfakes, disinformation campaigns, and poisoned models could compromise AI diplomacy tools, causing them to recommend flawed strategies or misinterpret adversarial intent. Without robust AI security measures and human oversight, the risks of error or exploitation can be considerable.
Conclusion
As a small and highly connected nation, Singapore depends on strong diplomatic ties for trade, investment and security. AI can offer powerful capabilities for conflict prevention and resolution, helping Singapore navigate this uncertain geopolitical landscape effectively.
In particular, AI-augmented diplomacy can support the management of longstanding bilateral issues, such as those involving natural resources or territorial boundaries, by modelling the consequences of various negotiation strategies. This equips diplomats with better foresight and adaptability at the negotiating table, promoting mediation and conflict resolution. Additionally, it can support real-time monitoring of regional instability – from civil unrest to maritime threats – bolstering Singapore’s strategic planning and civic preparedness.
As David Lammy pointed out, “the question before us is not whether AI will shape foreign policy. It is who will shape it and how”. AI will reshape the diplomatic landscape. The focus going forward should be on how Singapore can leverage AI to strengthen its diplomatic agility and safeguard regional stability in an increasingly polarised world.
About the Authors
Asha Hemrajani is a Senior Fellow at the Centre of Excellence for National Security (CENS) at S. Rajaratnam School of International Studies (RSIS), Nanyang Technological University (NTU). She is also a practising conflict resolution mediator. Rick Tan is a lieutenant in the Navy of the Singapore Armed Forces.
SYNOPSIS
Artificial Intelligence (AI) is set to become a game-changer in the field of diplomacy. From crisis response to conflict prevention and resolution, AI holds significant potential to enhance core diplomatic functions. However, its limitations necessitate a cautious, human-centred approach. For small states like Singapore, AI presents a valuable tool to enhance strategic foresight and resilience amid growing geopolitical volatility.
COMMENTARY
On his maiden visit to Singapore as the UK’s Foreign Secretary, David Lammy outlined his government’s vision for diplomacy in the digital age, stating that Artificial Intelligence (AI) will “transform how diplomacy is conducted”, because, by 2025, “diplomacy needs machine speed”. He outlined a future in which AI will revolutionise how nations project influence, anticipate threats, and forge alliances in an increasingly volatile world. He asserted that AI would transform how diplomacy is conducted, which will always have a human touch at its heart.
For millennia, diplomacy has been a fundamentally human endeavour, crafted through dialogue, negotiation, and the navigation of competing interests. Today, AI is ushering diplomacy into a new era.
AI enables rapid data processing, predictive analytics, and real-time decision support. These present opportunities for enhancing a variety of diplomatic functions, including crisis response, conflict prevention, and conflict resolution.
However, its limitations in diplomacy remain apparent, especially in its algorithmic biases and its inability to replicate the role of human intuition in tense settings. A measured approach is needed to integrate AI capabilities into the diplomatic domain.
Data-Driven Diplomacy
AI’s most significant comparative advantage is its ability to process vast datasets at speeds far beyond human capabilities. This enables the automation of time-consuming administrative tasks, such as document drafting and schedule management, freeing up diplomats to focus on strategic and relational functions.
As former US Secretary of State Antony Blinken noted, AI tools deployed in the State Department have significantly reduced the burden of manual tasks, such as summarising and translating diplomatic cables.
AI-powered predictive models are also used to predict conflicts or instability among countries. Researchers at Munster Technological University developed a model combining Bayesian Deep Learning and Random Forest techniques to forecast the stability of any country by analysing its historical and qualitative data, including traditional and social media. Predictions from diverse data inputs are generated, serving as early warning mechanisms for policymakers to engage in pre-emptive diplomacy or mediation.
Negotiation and Communication
AI models can also transform the architecture of negotiations, especially in designing peace agreements and engineering communication between parties. Trained on data from strategy games, media analyses, and historical peace agreements, the model can generate draft proposals for conflict resolution. Users input preferences across four areas: territory, security, justice, and economic arrangements. AI then evaluates the negotiability of each component for major stakeholders in the conflict.
In Libya’s peace process, the UN partnered with the AI platform Remesh to host large-scale, anonymised digital consultations with up to 1,000 participants per session. Participants could respond to key questions on governance, security, and elections, with the platform identifying areas of consensus and disagreement in real-time.
Accessible via low-bandwidth mobile devices and supporting Arabic dialects, the tool helped to shape the post-conflict roadmap, demonstrating how AI can enhance inclusion and ground peace talks in public input.
Language barriers have long complicated international negotiations, forcing diplomats to rely on interpreters who must navigate nuance, tone, and cultural subtext. AI-powered translation tools, such as DeepL, which offer real-time speech-to-text conversion, can facilitate smoother multilateral discussions.
These platforms also incorporate natural language processing, which can detect rhetorical shifts, tone, or implicit cues in speech. This can provide negotiators with deeper insights into their counterparts’ positions, particularly valuable in high-stakes settings where miscommunication can risk escalating conflict.
Crisis Management
AI can assist crisis management by analysing satellite imagery, social media trends, and news reports to predict displacement patterns and optimise manpower and resource distribution. The World Bank, for instance, has developed an AI-driven model to predict refugee influxes from South Sudan and the Democratic Republic of Congo into Uganda, drawing on hard data and online information related to the countries. Such a model can enable public services to be scaled up months before the refugees’ arrival, while also allowing for refugee integration into national services and local planning processes.
AI can also power intelligent Decision Support Systems (DSS). These systems aim to simulate optimal responses and propose best-practice solutions during high-stakes scenarios. However, DSS technologies are still in their early stages and present significant limitations. For example, the Futures Lab found that LLaMA, a generative AI model, opted for the use of force in 45 per cent of simulations. This highlights the need for further training and calibration to ensure such systems promote conflict de-escalation. Ultimately, DSS must be able to translate natural language inputs into reliable outputs that align more closely with human intent and ethical boundaries.
Simulation and Scenario Testing
Technologies like Extended Reality (XR) and digital twins, virtual models of real-world environments, allow diplomats to simulate complex negotiations. These tools test policy outcomes in risk-free settings, identifying unintended consequences before commitments are made. For example, a digital twin of a maritime dispute can simulate the outcomes of actions such as naval blockades or no-fly zones, projecting their economic and political ripple effects.
One such initiative, North Star, developed by tech start-up Anadyr Horizon, assigns probabilities of conflict or escalation to a wide range of decisions, helping decision-makers visualise risk in real time. These simulations not only enhance crisis preparedness but also support long-term scenario planning, enabling states to rehearse responses to hypothetical crises and anticipate the secondary effects of major geopolitical developments.
Limitations and Concerns
Despite the potential of AI systems in diplomacy, Professor Stuart Russell of the University of California, Berkeley, cautions against ”AI solutionism”, the view that AI is a universal problem solver. Geopolitical issues often involve deeply complex and sensitive human conflicts, where diplomacy still relies on trust and empathy, qualities that are uniquely produced through human-led interaction.
The 1970s US-China rapprochement, for instance, relied heavily on the personal rapport between Henry Kissinger and Zhou Enlai, and no algorithm could replicate the subtlety of their backchannel negotiations. While AI can provide data-driven insights, the final judgment must remain with diplomats attuned to historical context and emotional undercurrents.
Additionally, AI tools are also vulnerable to adversarial manipulation. Deepfakes, disinformation campaigns, and poisoned models could compromise AI diplomacy tools, causing them to recommend flawed strategies or misinterpret adversarial intent. Without robust AI security measures and human oversight, the risks of error or exploitation can be considerable.
Conclusion
As a small and highly connected nation, Singapore depends on strong diplomatic ties for trade, investment and security. AI can offer powerful capabilities for conflict prevention and resolution, helping Singapore navigate this uncertain geopolitical landscape effectively.
In particular, AI-augmented diplomacy can support the management of longstanding bilateral issues, such as those involving natural resources or territorial boundaries, by modelling the consequences of various negotiation strategies. This equips diplomats with better foresight and adaptability at the negotiating table, promoting mediation and conflict resolution. Additionally, it can support real-time monitoring of regional instability – from civil unrest to maritime threats – bolstering Singapore’s strategic planning and civic preparedness.
As David Lammy pointed out, “the question before us is not whether AI will shape foreign policy. It is who will shape it and how”. AI will reshape the diplomatic landscape. The focus going forward should be on how Singapore can leverage AI to strengthen its diplomatic agility and safeguard regional stability in an increasingly polarised world.
About the Authors
Asha Hemrajani is a Senior Fellow at the Centre of Excellence for National Security (CENS) at S. Rajaratnam School of International Studies (RSIS), Nanyang Technological University (NTU). She is also a practising conflict resolution mediator. Rick Tan is a lieutenant in the Navy of the Singapore Armed Forces.
