The Labs
BTI’s work is organised into standing research groups and a flagship program.
Capabilities R&D
You can’t measure the future, but you can measure a difference. Lab A runs the same task across a span of models, deliberately including older and weaker ones, and reports how a result is affected during capability changes.
Studies run here
Cyber Operations
Lab C runs the synthetic range. Security tasks are conducted the established way to set a rigorous baseline, then conducted again by AI with identical access. Everything runs against environments we build and own, so the true state of the target is known by construction and every number can be checked against it.
Studies run here
Project BLUE LIGHT
AI raises the capabilities of a lone wolf attacker and of a defender. The gap between those two curves is a reflection of both the current state of AI capabilities and security procedures. BLUE LIGHT measures both sides on the same environment and reports which improves faster, and where they cross.
