DSIT Launches £15m Fund for AI Security Research
The UK's Department for Science, Innovation and Technology (DSIT) has unveiled a new research fund exceeding £15 million, aimed at studying the behaviour and control of artificial intelligence. On 30 July, DSIT announced that its AI Security Institute will lead efforts focused on AI alignment, ensuring AI systems act in accordance with their intended design.
The initiative, called the Alignment Project, brings together an international coalition including the Canadian AI Safety Institute, Amazon Web Services, US-based AI company Anthropic, UK Research and Innovation, and the Advanced Research and Invention Agency.
Peter Kyle, the science and technology secretary, stressed the importance of advancing research in AI behaviour as these systems evolve at increasing speeds. He highlighted the need to guarantee that AI technologies operate in humanity’s best interests.
Charlotte Deane, executive chair of the Engineering and Physical Sciences Research Council (EPSRC), noted the partnership's role in bridging fundamental AI science with practical alignment challenges. The collaboration is expected to strengthen the UK's AI ecosystem by connecting discovery science with real-world applications.
Focus on Safe and Responsive AI Systems
The Alignment Project will fund research to ensure AI remains under human oversight and can detect and mitigate behaviours that might pose societal risks. Researchers from fields such as computer science and cognitive science can apply for grants of up to £1 million.
Additionally, participants will gain access to computing resources from Anthropic and Amazon Web Services, enabling experiments that typically exceed the capacity of academic institutions.
Addressing a Critical Challenge in AI Development
Geoffrey Irving, chief scientist at the AI Security Institute, emphasized that AI alignment is a pressing and underfunded issue. He pointed out that progress in this area is not keeping pace with the rapid advancements in AI technologies.
The project also benefits from private sector investment aimed at accelerating commercial solutions for AI alignment. Its advisory board includes esteemed figures such as Turing Award winners Yoshua Bengio and Shafi Goldwasser.
AI is central to current government strategies for economic growth. DSIT believes that the Alignment Project can help overcome one of the main barriers to safe AI adoption: building public trust.
According to Charlotte Deane, EPSRC remains committed to supporting pioneering research in AI safety. By coordinating efforts with the AI Security Institute and other partners, the UK is reinforcing the coherence and strength of its AI ecosystem.
She added, “Together, we will ensure that the UK’s leadership in artificial intelligence delivers both breakthrough innovation and real-world benefits.”
Your membership also unlocks:
