Under the agreement, the UK will gain access to a vast repository of real-world conflict data collected by Ukraine, including drone footage, electronic warfare signatures, and ground sensor readings. The data will be used to develop AI models that can detect, predict, and respond to physical threats such as sabotage, surveillance, or coordinated attacks on critical national infrastructure.
Private companies are also expected to be granted access to the dataset to build commercial security systems, according to officials familiar with the deal. The arrangement marks a significant deepening of defence and technology ties between London and Kyiv, building on previous cooperation in areas such as cyber defence and arms supplies.
The UK government has emphasised that the AI systems will be used for protective, not offensive, purposes. However, the inclusion of protester activity alongside hostile state actors has drawn attention, raising questions about the scope of surveillance and the potential for the technology to be used in domestic policing.
Sources in London said the data would be anonymised to protect operational security and that all models would undergo rigorous testing before deployment. Ukraine has been using its Avengers AI lab for over a year to process vast amounts of battlefield information, and its models have proven effective in identifying emerging threats in real time.
The deal is part of a broader UK strategy to accelerate the adoption of AI in defence and national security. Other initiatives include a new AI safety institute focused on military applications and increased investment in autonomous systems.