George Balston

Working to ensure democracies can adopt powerful AI systems safely and competitively.

I am Co-Founder at AVERI, a non-profit working to make third-party auditing of frontier AI effective and universal.

Previously, I served as Co-Director for Defence & National Security at The Alan Turing Institute, where I founded CETaS and led a research programme on emerging technology and national security.

Before that, I worked for the U.K. Government researching, developing, and deploying new analytical capabilities.

A few things I've worked on recently

  1. 01
    Double-Blind AI Evaluations

    Piloting the world’s first double-blind evaluation of a proprietary AI model, using secure enclaves to keep both the model weights and the benchmark confidential.

    With Google DeepMind OpenMined MLCommons

    27 Aug 2026 storage.googleapis.com/deepmind-media/DeepMind.com/Blog/piloting-the-worlds-first-double-blind-ai-evaluations/double-blind-evaluations-technical-report.pdf

  2. 02
    AI Sovereignty and National Security

    Identifying the UK’s dimensions of control with respect to AI sovereignty in the context of national security

    With CETaS

    20 July 2026 cetas.turing.ac.uk/publications/ai-sovereignty-and-national-security-identifying-uks-dimensions-control

  3. 03
    State of the Art Evaluations for Loss of Control

    Analysing the need for deeper white box access to maintain state of the art evaluations for loss of control threats

    With Apollo Research

    20 May 2026 apolloresearch.ai/governance/the-need-for-deeper-white-box-access-to-maintain-state-of-the-art-evaluations-for-loss-of-control-threats

  4. 04
    Secure Access to Frontier AI Taskforce

    Exploring the cyber security, information security, and model-level risks associated with third-party access to frontier AI models.

    With RUSI Google

    12 May 2026 rusi.org/explore-our-research/projects/secure-access-frontier-ai-taskforce

  5. 05
    Launching AVERI

    A non profit with an aim to make third party auditing of frontier AI effective and universal.

    15 Jan 2026 x.com/AVERIorg

  6. 06
    Frontier AI Auditing

    Toward Rigorous Third-Party Assessment of Safety and Security Practices at Leading AI Companies.

    15 Jan 2026 averi.org/ourwork/frontier-ai-auditing

  7. 07
    AEF-1

    Minimum Operating Conditions for Independent Third Party AI Evaluations.

    With Transluce

    4 Dec 2025 aievaluatorforum.org/initiatives/minimum-operating-conditions