About — the residency
We are founders, engineers, researchers, and policymakers who want to secure frontier AI.
The Frontier AI Security Residency (FASR) is an 8-week, full-time residency dedicated to the hardware and cyber security of frontier AI: we want to push forward ambitious projects for securing frontier models, and especially support domain experts work directly with orgs in this space.
We support engineers, entrepreneurs and researchers to take on concrete, high-leverage security problems amid rapidly advancing AI capabilities, working beside the people tackling them at the frontier. The residency is designed to meet this reality.
Run by ERA, with the Oxford Hardware AI Governance Lab and Heron AI Security Initiative.
Team — core & research managers
Partnership — who runs the residency
The Frontier AI Security Residency is run by ERA, in partnership with Heron and the Oxford Hardware AI Governance Lab.
ERA runs one of the world's largest, longest-running AI safety and governance talent programmes in the world. We are a talent organisation founded by Cambridge alumni, and based at Cambridge, UK, with strong ties to the University of Cambridge. We support researchers and entrepreneurs working to mitigate risks from frontier AI. So far, ERA's mission has been largely manifested through our annual ERA:AI fellowship programme.
In 2025, we ran a Summer ERA:AI Research Fellowship, hosted 30+ speakers & workshops over the 8 weeks, ran a Technical AI Governance Forum, and co-organised the 2025 Vegas AI Security Forum. You can see some of our past research here. Over the past 5 years, we have supported over 120 early-career researchers from 10+ countries through our research fellowship and conferences, leading to high counterfactual impact on their careers. We provide our fellows with mentorship from organisations such as UK AISI, CAISI, RAND, GovAI, Google DeepMind etc., and our alumni have gone on to lead work at impactful institutions in this space.
The Hardware AI Governance Lab (HAIGL) is a newly established interdisciplinary initiative based at the University of Oxford. It brings together expertise from both computer hardware architecture and AI governance to explore how technical and policy-oriented approaches can be combined to address emerging challenges in advanced AI systems. At its core, HAIGL seeks to investigate innovative hardware-level solutions that can support the governance of AI. Rather than focusing solely on software or regulatory frameworks, the lab emphasizes the potential of hardware architecture as a foundational layer for embedding oversight, verification, and control mechanisms. This approach is particularly relevant in light of increasing geopolitical tensions, as well as the growing complexity and strategic importance of the global semiconductor supply chain.
Heron bridges the gap between the frontier AI models and the level of cybersecurity they need, to minimize risks to society. A non-profit based in Tel Aviv and launched in December 2024, Heron seeks to build a community of world-class experts who want to shape the future of technology while advancing their careers. Their focus is practical and impact-driven: enabling cybersecurity experts to work where their skills matter most.
Heron runs events, curated meetups, and personalized 1-1 networking that connect professionals to high-leverage opportunities in AI security. They develop research and learning programs that help experienced practitioners ramp quickly on core technical workstreams.