The UK government has unveiled a new £15m global initiative aimed at solving one of AI’s toughest technical challenges – ensuring that advanced AI systems act in ways that align with human values.
The Alignment Project, led by the UK’s AI Safety Institute, will fund academic and non-profit research teams with grants of up to £1m.
Industry heavyweights including Amazon Web Services (AWS) and Anthropic are backing the scheme, offering cloud compute support to accelerate research into AI safety, transparency and control mechanisms.
“Advanced AI systems are already exceeding human performance in some areas, so it’s crucial we’re driving forward research to ensure this transformative technology is behaving in our interests,” said Science, Innovation and Technology Secretary Peter Kyle.
“AI alignment is all geared towards making systems behave as we want them to, so they are always acting in our best interests.
“The responsible development of AI needs a co-ordinated global approach, and this fund will help us make AI more reliable, more trustworthy, and more capable of delivering the growth, better public services, and high-skilled jobs that drive our Plan for Change.”
The initiative builds on momentum from the UK’s 2023 AI Safety Summit, reinforcing its ambition to lead on global safety standards and technical safeguards in frontier AI – powerful models capable of autonomous reasoning, decision-making, and code generation.
However, experts remain sceptical about the UK’s influence, given the voluntary nature of the programme and limited access to proprietary systems developed by private companies such as OpenAI and Google DeepMind.
The AI Safety Institute has struggled to fully test these models due to a lack of transparency around training data and system internals.
Critics argue that without enforcement powers or independent auditing rights, the government’s efforts may remain largely symbolic.
Others have stressed that a true leadership position in AI safety will also require massive investment in workforce development.
Sheila Flavell CBE, COO of FDM Group, commented: “AI has the potential to transform every sector of our economy, but real leadership on the global stage will depend on our ability to equip people with the right skills. Building a workforce fluent in AI, not just in using tools, but in understanding their capabilities and limits, is essential.
“We must invest in practical, inclusive training pathways that open doors into tech careers for people from all backgrounds. AI levels the playing field and opens up new opportunities, especially for underrepresented backgrounds within the tech sector, from graduates to returners to ex-forces, who can launch a career with AI as a driver.
“Experiential learning, simulating real-world scenarios that businesses face every day, can provide comprehensive training at speed, helping people to build the AI skills to solve industry challenges quickly.
“Nurturing future talent is crucial and sustained collaboration between government, industry, and education will have a lasting impact, allowing the UK to lead not just in innovation, but in inclusive, human-centred AI adoption.”