AI in International Security and Conflict

AI in International Security and Conflict

Title history
  • AI in International Security and Conflict — observed on 20261107-0827 UTC (YouTube)
🎙 Dickey Center for International Understanding 👥 4K 📅 November 7, 2026 ⏱ 60 min 👁 61 📄 expert opinion 🧭 2026-11-07
Available in: English (current) Français

Keywords

artificial intelligencewarfaredronespolicyautonomous weapons

Summary

The panel, featuring Sasha Baker from OpenAI and James Patton Rogers from King’s College London, moderated by Ambassador Erica Barks-Ruggles, examines how AI is transforming international security and conflict. Sasha Baker highlights two fundamental changes: increased speed and scale of operations, which challenge existing military and governmental bureaucracies. James Rogers focuses on the proliferation of drones and autonomous weapons, noting that smaller states and non-state actors can now leverage cheap technologies to counter great powers, as seen in Ukraine and the Middle East. They discuss the ethical and regulatory challenges posed by open-source models and the difficulty of controlling access. The panel also considers future scenarios: Rogers worries about low-tech autonomous systems that lack discrimination, while Baker predicts limited adoption in Western militaries due to bureaucratic inertia, unless existential threats force rapid change. The audience raises questions on regulation and secrecy, comparing AI to nuclear proliferation. The discussion emphasizes the need for new partnerships between governments and the private sector, and for adapting international norms to this fast-evolving domain.

169 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of this panel lies in the complementary perspectives: Baker’s insider view from government and industry, and Rogers’ academic research on drone warfare. They effectively argue that AI’s speed and scale are reshaping the balance of power, making advanced capabilities accessible to a wider range of actors. Their arguments are well-structured, with concrete examples (the PHP vulnerability, Ukrainian drone innovation, Iranian attacks) grounding the discussion. However, the claims are not backed by statistical data or peer-reviewed studies, and some predictions remain speculative. The exchange between Baker and Rogers on the adoption rate of AI in Western militaries demonstrates a healthy debate, though neither provides empirical evidence for their forecasts.

Scientific Rigor, Source Quality, Title Accuracy

The discussion is rigorous in its use of expert knowledge and real-world cases, but does not cite specific sources or studies. Panelists mention the ‘What Must Remain Human?’ conference and OpenAI’s role in cybersecurity, but no scientific publications. The title accurately matches the content, which focuses on AI’s intersection with international security and conflict. The panel’s credibility rests on the speakers’ extensive experience, but the lack of verifiable citations limits its scientific rigor. The audience comments are not analyzed as they were not provided.

209 words

Title / Content Match

Titre accurately reflects the subject matter: the panel discusses AI's impact on warfare, security, and policy.

Quality & Reliability

7/10

Panel of recognized experts in national security and drone warfare, moderated by a former ambassador. Discussion based on experience and analysis, but relies on informed opinion rather than peer-reviewed evidence.

Key Moments

Contribution & Novelties

This panel provides a timely and informed discussion on AI’s role in warfare, combining high-level policy insight (Baker) with frontline research on drone proliferation (Rogers). It highlights the tension between advanced AI safety initiatives and the unstoppable spread of open-source components, offering a nuanced view of future threats. The debate on adoption inertia vs. crisis-driven innovation is particularly valuable for policymakers.

Pour aller plus loin :

99 words

Radar Profile

The radar profile shows moderate to high scores in information quality and reliability, with lower technical level and quantity reflecting a discussion-oriented, non-quantitative approach. The emphasis on expert opinion without empirical data leads to a balanced but slightly below-maximum score.

Reliability 7/10