Research
Five areas of applied security research. Each one feeds a service you can commission.
How models and AI applications are attacked, and how machine learning detects threats.
We study attacks to build defences that hold up in deployed systems, not only in the lab. The work goes straight into how we test AI applications for clients.
Feeds → AI security and assurance · Secure software and AI development
Smart card security, trusted execution environments, smartphone and IoT security.
Hardware is where trust starts. We look at devices from the chip up, which is why our device reviews find what software-only reviews miss.
Feeds → Cybersecurity consulting · Applied research and prototyping
Designing and implementing security protocols that do what they claim.
Most protocol failures happen in implementation, not on paper. We design and review both.
Feeds → Secure software and AI development · Applied research and prototyping
User-centred privacy, and who owns and controls data across devices and clouds.
Privacy is a property you design for, not a policy you write afterwards. We help teams build it in.
Feeds → Cybersecurity consulting · AI security and assurance
Blockchain, drones and autonomous vehicles.
New technology brings old mistakes in new forms. We find them early, while they are cheap to fix.
Feeds → Applied research and prototyping
The record
Titles, venues and years of our research outputs, grouped by area.
See selected workAcross the team
Collaborations and funding
We collaborate on joint research, grant bids and prototypes with academic and industry partners. Partners and funders are listed here only with their permission.
TBC · partners and funders who have agreed to be named
Next step
Joint bids, feasibility studies or a hard question you want answered.