…
…
Tech & Science
Herz — Tech & Science Desk · · 30s summary · 3 min read
London and Kyiv concluded an agreement on August 24, 2026, to train artificial intelligence models using Ukrainian battlefield data. The systems are intended to protect British military sites and critical infrastructure from protesters and hostile states. An initial pilot at a UK defence site will combine buried fibre-optic cables with AI-enhanced sensors designed to distinguish movements by people and vehicles. Private companies will require Ministry of Defence approval to access the data through a secure platform. Sintela, Mind Foundry and Skyral have already received approval for pilot projects.
London and Kyiv concluded an agreement on August 24, 2026, to train artificial intelligence models using Ukrainian operational data. According to The Guardian, the tools are intended to protect British military sites and critical infrastructure.
The systems will be designed to identify protesters or an attack by a hostile state. Presented as the first agreement of its kind in the UK, the system will be tested at a British defence site.
The pilot will use AI-enhanced sensors integrated into buried fibre-optic cables. They are intended to distinguish different movements by people or vehicles around sensitive sites.
Private companies will be able to access data from Ukraine’s Avengers AI lab through a secure platform. Every request will require approval from the UK Ministry of Defence.
The data includes drone operations, imagery from strike missions and Ukrainian flight profiles. It may also cover Russian attempts to sabotage critical infrastructure.
The military dataset covers four years. It provides a deeper record than the mainly open-source data previously used to train models outside Ukraine.
Sintela, Mind Foundry and Skyral have already received approval for pilot projects. All three are British companies working in artificial intelligence.
Bristol-based Sintela, which specialises in fibre-optic sensors, has also signed a contract with the Trump administration to monitor the US-Mexico border. The contract is worth $35 million, equivalent to £26 million.
If the technology proves effective, the government is considering deployment at airports, prisons, rail networks and energy infrastructure. The electricity grid and military bases are also among the sites identified.
The government ultimately hopes the models will rapidly anticipate threats or tensions caused by hostile actors against railways, the electricity grid or military bases.
The illustrations in this article are generated by artificial intelligence.
No comments yet. Be the first to react.
The partnership could also support projects involving new artificial intelligence chips for drones, robotics and automated weapons systems.
Sources at the UK Ministry of Defence said the system is explicitly intended to detect actions by protesters against sensitive sites. Its purpose is therefore not limited to attacks by hostile states.
In June 2026, activists entered RAF Brize Norton in Oxfordshire, the Royal Air Force’s largest base. They arrived on scooters, spray-painted two military aircraft and left without being arrested.
The controversial proscription of Palestine Action followed the intrusion. Founded in 2020, Palestine Action is a British pro-Palestinian protest network that targets companies in the military sector, among others.
The group said it had targeted RAF Voyagers, the Royal Air Force name for Airbus A330 MRTT aircraft used for transport and air-to-air refuelling. It said its aim was to disrupt what it described as British participation in crimes in the Middle East.
Steven Murdoch, a professor of computer science at University College London, noted that sensors embedded in fibre-optic cables have been detecting movement for decades.
He said it was uncertain what Ukrainian data would add to that function, while arguing that the information could plausibly improve detection of newer threats, including drones.
Andy Burnham said combining British artificial intelligence expertise with Ukrainian data would accelerate the development of new security capabilities.
Those capabilities are intended to strengthen national security and protect critical infrastructure. Wes Streeting said Ukraine was pushing the boundaries of defence innovation as it resisted the Russian invasion.
It is intended to train artificial intelligence models that protect British military sites and critical infrastructure.
The partnership covers drone operations, imagery from strike missions and Ukrainian flight profiles collected over four years.
It will use AI-enhanced sensors integrated into buried fibre-optic cables.
Sintela, Mind Foundry and Skyral have received approval for pilot projects.
The government has identified airports, prisons, rail networks, energy infrastructure, the electricity grid and military bases.