Hi all,
I am sharing a technical update from our team at the Alan Turing Institute, part of our ongoing work under the Defence and National Security Grand Challenge.
In this post, Airgapped No More – May 2025 Snapshot, we outline how we use AI to advance cyber defence in critical national infrastructure environments. A few things we cover:
Why traditional IT security tools fail in critical infrastructure
How our system fakes entire OT networks using LLM-based responders
The architecture for our protocol-aware deception at Layer 2
This is part of our broader mission to shift cyber risk asymmetry in favour of defenders, especially those protecting fragile, under-resourced infrastructure. If you're working on OT security, malware analysis, or AI safety, we'd love your feedback or to collaborate.
Read the update: https://airgapped.substack.com/p/update-may-2025-technical-snapshot
Code access: https://github.com/alan-turing-institute/smoke-and-mirrors
Kind Regards,
Vasilios Mavroudis