Lisa Soder
Tagline:Hi! I am Lisa and I work on European AI policy and building out the AI Accountability Infrastructure!
Vienna, Austria
About Me
I am a Senior Policy Researcher at interface, a pan-European technology policy think tank, and a Research Affiliate with RAND’s Technology and Security Policy Center. My work examines how to make AI governance both safe and beneficial for society, with a particular focus on designing technical governance measures—such as risk assessment and auditing—and exploring the geopolitical dimensions of AI, including the role of China-EU relations.
Prior to joining Interface, I served as a Summer Research Fellow at the Centre for the Governance of AI, worked as a Strategy Consultant at the Boston Consulting Group, and led a global health NGO in Ethiopia. I hold a Master’s degree from the London School of Economics, where I focused on comparative transatlantic approaches to technology regulation and competition law.
Outside of work, you can often find me in the backcountry—ski touring, hiking, or cycling!
Contact: You can reach me under lisasoderhoehne [at] gmail.com
Publications
Position: Technical Research and Talent is Needed for Effective AI Governance
Conference PaperPublisher:Forty-first International Conference on Machine LearningDate:2025Authors:Anka Reuel*Lisa Soder*Benjamin BucknallTrond Arne Undheim
An Autonomy-Based Classification: AI Agents, Liability and Learnings from Automated Vehicles Act
Conference PaperPublisher:Workshop on Socially Responsible Language Modelling ResearchDate:2025Authors:Lisa SoderJulia SmakmanConnor DunlopWeiwei PanSiddharth SwaroopNoam KoltPosition: AI Agents & Liability–Mapping Insights from ML and HCI Research to Policy
Conference PaperPublisher:Workshop on Socially Responsible Language Modelling ResearchDate:2025Authors:Weiwei PanSiddharth SwaroopConnor DunlopJulia SmakmanLisa SoderGPAI Evaluations Standards Taskforce: Towards Effective AI Governance
Journal ArticlePublisher:arXiv preprint arXiv:2411.13808Date:2024Authors:Patricia PaskovLukas BerglundEverett SmithLisa SoderOpen Problems in Technical AI Governance
ReportDate:2024Authors:Anka ReuelBen BucknallStephen CasperTim FistLisa SoderOnni AarneLewis HammondLujain IbrahimAlan ChanPeter WillsMarkus AnderljungBen GarfinkelLennart HeimAndrew TraskGabriel MukobiRylan SchaefferMauricio BakerSara HookerIrene SolaimanAlexandra Sasha LuccioniNitarshan RajkumarNicolas MoësJeffrey LadishNeel GuhaJessica NewmanYoshua BengioTobin SouthAlex PentlandSanmi KoyejoMykel J KochenderferRobert TragerMore than Marketing? On the Information Value of AI Benchmarks for Practitioners
Journal ArticlePublisher:arXiv preprint arXiv:2412.05520Date:2024Authors:Amelia HardyAnka ReuelKiana Jafari MeimandiLisa SoderAllie GriffithDylan M AsmarSanmi KoyejoMichael S BernsteinMykel J KochenderferTowards Publicly Accountable Frontier LLMs: Building an external scrutiny ecosystem under the ASPIRE framework
Journal ArticlePublisher:arXiv preprint arXiv:2311.14711Date:2023Authors:Markus AnderljungEverett Thornton SmithJoe O'BrienLisa SoderBenjamin BucknallEmma BluemkeJonas SchuettRobert TragerLacey StrahmRumman Chowdhury
Talks
Voluntary AI Commitments: Paris AI Action Summit
Date: Feb 2025
Location: Paris .
Description:I joined a panel at the AI Action Summit to discuss progress on frameworks like the Frontier AI Safety Commitments (FAISC) and the Hiroshima AI Process (HAIP)—examining how these efforts relate to each other, what has been achieved so far, and the challenges of turning early governance steps into genuine accountability.
Other panelists included Takuo Imagawa (Vice-Minister, Ministry of Interior and Communications, Japan), Sara Hooker (VP of Research, Cohere AI), John McDermid (Professor, Department of Computer Science, University of York), Peter Sarlin (CEO, AMD Silo AI). The disucssion was moderated by Audrey Plonk (Deputy Director, Directorate for Science, Technology and Innovation, OECD) moderated the discussion.
Sino-German Track 2 Dialogue
Date: Dec 2024
Location: Beijing & Shanghai .
Description:I joined a Sino-German Track 2 dialogue in Beijing and Shanghai. The agenda included topics such as economic security, international AI governance, and the regulation of emerging technologies.
In the Media
AI Action Summit: Stresstest für globale KI-Regeln
Issue date: Feb 2025,
Issued by: Tagesspiegel .
AI Action summit in Paris: A stress test for global AI governance
Issue date: Feb 2025,
Issued by: Context (Thomson Reuters Foundation) .
AI: US under Trump and Europe choose diverging paths
Issue date: Jan 2025,
Issued by: Deutsche Welle (Janosch Delcker) .