All research

Signals in the Noise: Open Source Intelligence (OSINT) for AI Loss of Control Detection

Authors

Sarah Bollinger

Nada Aboserie

Amanda Coakley

Ian Lee

Taysir Mathlouthi

Research team leader

Sarah Bollinger

expert partner

Tommy Shaffer Shane (CLTR - Centre for Long-Term Resilience)

AI Governance Taskforce

Winter 2026

Full Paper

Summary

Policy brief

This paper applies open-source intelligence (OSINT) and cyber threat intelligence (CTI) methodologies to the problem of detecting AI systems operating outside human control. Drawing on a cross-disciplinary literature review and 14 semi-structured expert interviews conducted under Chatham House Rule, the paper develops two threat models, identifies a range of observable traces, and proposes an institutional architecture for monitoring. The research finds that OSINT-based detection of loss of control is partially feasible and worth building now. Three detection vectors emerge as highest priority: transcript-based collection of user-reported AI behaviour; infrastructure correlation for unexpected external connections or replication; and output analysis for capability concealment. The paper argues for a dedicated, federated international monitoring capability anchored in OSINT methods and independent of frontier AI developers, and identifies sustained non-industry funding as the highest-leverage structural intervention available.

Expert Partner: Tommy Shaffer Shane (CLTR - Centre for Long-Term Resilience)

SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6735558

Alumni

Meet the authors

(Research Team Lead)

Sarah Bollinger

Sarah Bollinger is a Arcadia Senior Research Fellow at the Center for AI Risk Management & Alignment (CARMA) and Research Team Lead at Arcadia Impact, examining the intersection of geopolitical risk analysis and emerging technology policy.

With over a decade of experience spanning conflict analysis, strategic intelligence, and geopolitical risk assessment across humanitarian and private industry contexts, Sarah examines how political economy, institutional structures, and sociotechnical dynamics shape AI governance outcomes. She specialises in translating complex technical concepts for policymakers and bridging disciplinary divides between risk analysis and policy implementation.

Nada Aboserie

Nada is a technology policy expert specialising in AI governance, data sovereignty, and digital regulation across Europe, the Middle East, Africa, and Asia Pacific.

As Deputy Director of AI at Access Partnership, she shapes regulatory strategy for Fortune 100 companies in complex high-stakes environments.

Before Access Partnership, she led government affairs programs at Microsoft across 24 markets. She holds CIPP-E and AIGP certifications and has called nine countries home.

Amanda Coakley

Amanda is a tech policy and global affairs strategist focused on AI governance, frontier technologies, and geopolitical risk.

She is an AI safety fellow at the AI Safety Initiative at Georgia Tech, a research associate with Arcadia Impact’s AI Governance Taskforce, and a research group member at the Centre for AI and Digital Policy, contributing to international AI governance analysis.

She is also a columnist at World Politics Review writing on European politics and international affairs.

Ian Lee

Ian is a geopolitical intelligence consultant with a strong foundation in OSINT, AI security analysis, and technological geopolitics, specifically the dynamics between APAC and the western world.

He applies his threat intelligence expertise to research behavioral signatures, focusing on distinguishing autonomous AI operations from human-controlled malicious use of AI andleverages his geopolitical and OSINT skills to understand how AI may be misused to create inequality and catastrophic risks.

Ian is dedicated to mitigating adversarial AI risks, addressing AI hegemony and technical lock-in, and ensuring equitable technology governance and security for all nations.

Taysir Mathlouthi

Taysir is an Advocacy and Policy expert specializing in the governance of AI and emerging technologies in conflict-affected settings.

Taysir is currently the Senior EU Advocacy Officer for Digital Rights and AI Governance at 7amleh, and a specialist in human rights and digital governance. Taysir has also worked with leading international and civil society organizations, including UNICEF and International Alert, where she contributed to policy research, advocacy campaigns, and public engagement on issues such as AI-driven disinformation, surveillance risks, and biased content moderation in conflict settings.

Her experience includes active engagement with EU and UN institutions, multi stakeholder consultations, and platform accountability efforts.

Alumni

Meet the authors

Sarah Bollinger

(Research Team Lead)

Sarah Bollinger is a Arcadia Senior Research Fellow at the Center for AI Risk Management & Alignment (CARMA) and Research Team Lead at Arcadia Impact, examining the intersection of geopolitical risk analysis and emerging technology policy.

With over a decade of experience spanning conflict analysis, strategic intelligence, and geopolitical risk assessment across humanitarian and private industry contexts, Sarah examines how political economy, institutional structures, and sociotechnical dynamics shape AI governance outcomes. She specialises in translating complex technical concepts for policymakers and bridging disciplinary divides between risk analysis and policy implementation.

Nada Aboserie

Nada is a technology policy expert specialising in AI governance, data sovereignty, and digital regulation across Europe, the Middle East, Africa, and Asia Pacific.

As Deputy Director of AI at Access Partnership, she shapes regulatory strategy for Fortune 100 companies in complex high-stakes environments.

Before Access Partnership, she led government affairs programs at Microsoft across 24 markets. She holds CIPP-E and AIGP certifications and has called nine countries home.

Amanda Coakley

Amanda is a tech policy and global affairs strategist focused on AI governance, frontier technologies, and geopolitical risk.

She is an AI safety fellow at the AI Safety Initiative at Georgia Tech, a research associate with Arcadia Impact’s AI Governance Taskforce, and a research group member at the Centre for AI and Digital Policy, contributing to international AI governance analysis.

She is also a columnist at World Politics Review writing on European politics and international affairs.

Ian Lee

Ian is a geopolitical intelligence consultant with a strong foundation in OSINT, AI security analysis, and technological geopolitics, specifically the dynamics between APAC and the western world.

He applies his threat intelligence expertise to research behavioral signatures, focusing on distinguishing autonomous AI operations from human-controlled malicious use of AI andleverages his geopolitical and OSINT skills to understand how AI may be misused to create inequality and catastrophic risks.

Ian is dedicated to mitigating adversarial AI risks, addressing AI hegemony and technical lock-in, and ensuring equitable technology governance and security for all nations.

Taysir Mathlouthi

Taysir is an Advocacy and Policy expert specializing in the governance of AI and emerging technologies in conflict-affected settings.

Taysir is currently the Senior EU Advocacy Officer for Digital Rights and AI Governance at 7amleh, and a specialist in human rights and digital governance. Taysir has also worked with leading international and civil society organizations, including UNICEF and International Alert, where she contributed to policy research, advocacy campaigns, and public engagement on issues such as AI-driven disinformation, surveillance risks, and biased content moderation in conflict settings.

Her experience includes active engagement with EU and UN institutions, multi stakeholder consultations, and platform accountability efforts.

Programme

AI Governance Taskforce

The AI Governance Taskforce is a career development programme for experienced professionals looking to transition careers into AI governance, focussed on reducing risks from advanced AI.
Participants work around existing commitments during our 12 week, remote, part-time cohorts, producing policy research in teams of 4, led by our Research Team Lead staff in partnership with recognised experts in the field. Teams write an academic-style paper and accompanying blog post to build knowledge, skills and work portfolios.

Learn more