The UK’s Ministry of Defence’s Defence Science and Technology Laboratory (DFTL) has signed a memorandum of understanding with Google Cloud that will see the duo work to accelerate the use of cloud and AI technology in the UK defence sector. The move will make Google’s AI models accessible for UK defence projects.
The collaboration will also include training from Google that’s tailored towards the defence sector, cross-sector technology transfer and the development of new ways of working.
MoD second parliamentary under-secretary of state, Paul Lincoln, revealed the agreement with Google during a talk at the AI Summit, a London Tech Week event, earlier today. He told delegates the MoD is becoming more agile in how it works with commercial partners: “We need to have more ambition in how we bring AI products to life, otherwise the cycle has moved on and we are already too late,” he said.
The Google agreement will focus on five key areas: acceleration of technology adoption, broadening the supply chain for MoD partners, support for training and upskilling, increasing cross-sector technology transfer and sharing new ways of working and learning from the tech sector.
DFTL will be able to access Google’s AI technologies, processes and people to understand how it delivers solutions to end users, and integrate that into its wider remit. This will also include integrating techniques used across civilian applications and other industries. The aim is to explore where tools developed for other purposes can help solve UK defence challenges.
“This won’t just be something with military benefits, but will also yield mainstream civilian benefits that enhance people’s lives,” declared Lincoln, speaking to an audience made up of AI industry figures including many from start-ups and scale-ups.
Embedding AI across defence
Similar announcements will follow in the coming weeks to “ensure we are in the best possible position to protect the UK from new and emergent threats, and to exploit the transformation opportunities presented by AI technologies,” he said.
“We are making efforts embedded in everything we do,” Lincoln said. “None of our AI work is undertaken in isolation. We work across the MoD, we work across Whitehall, we work across industry and we work across academia to ensure our work is not just ambitious but safe and crucially responsible.”
The work will include the Defence AI Centre (DAIC) formed last year by the MoD to accelerate its ability to harness the power of AI.
“We are delighted to have signed an MOU with DSTL on this important initiative,” said Helen Kelisky, UK and Ireland managing director at Google Cloud. “AI has potential to assist the ministry in a broad range of use cases, including cybersecurity, disaster response, and employee productivity. We look forward to helping DSTL in its commitment to using AI boldly and responsibly.”
Google has already signed agreements in the US and elsewhere on defence including with the US Defence Innovation Unit to provide cloud security management solutions, with the judicial branch in Peru to provide remote collaboration and with the Dusseldorf Fire Department for geospatial data access to help remote workers. The company says any AI work, regardless of the company or organisation, is done with its existing AI principles in mind. This includes rules around social benefit, avoiding bias, safety testing and accountability.
Lincoln told delegates security and risk mitigation would be central to any implementation, and use of AI in defence would be held to the same international humanitarian and ethical standards as other defence technologies. He warned that, while caution is needed, “our adversaries are also operating in this space,” and we “can’t afford to fall behind the pace in AI or leave it to adversaries to decide the shape of things to come.” He gave some use case examples, including protecting people in combat situations through more accurate simulations, rapid diagnosis of medical trauma and optimisations of deployment.