The Department for Science, Innovation and Technology (DSIT) has announced £100 million in start-up funding for an expert AI taskforce with plans for an early focus on public services.
It said the new body will work on the ensuring the sovereign capabilities and broad adoption of safe and reliable foundation models of AI systems trained on massive datasets.
This comes in response to the growing awareness of the potential but also concerns around technologies such as ChatGPT and Google Bard.
The taskforce will be modelled on the Covid-19 Vaccines Taskforce, include government and industry experts, and report directly to the prime minister and technology secretary. Its membership will be announced later in the summer, with Matt Clifford, chair of the Advanced Research and Innovation Agency, advising on its development.
It will aim to establish the UK as a world leader in the foundation models and their applications across the economy. This can make the country a global standard bearer for safe AI, DSIT said.
The taskforce will launch its first pilot projects targeting public services in the next six months.
Transformative impact
Science, Innovation and Technology Secretary Michelle Donelan, said: “Developed responsibly, cutting edge AI can have a transformative impact in nearly every industry. It can revolutionise the way we develop new medical treatments, tackle climate change and improve our public services, all while growing and future proofing our economy.
“We need to act now to seize the opportunities AI can offer us in the future. We’re backing our expert taskforce with the funding to make our ambitions for an AI enabled country a reality and keep the UK at the front of the pack in this emerging technology.
“To ensure such leadership, the greatest capability we can develop is in the safety and reliability of such systems. This will ensure that the public and business have the trust they need to confidently adopt this technology and fully realise its benefits. That is exactly what this taskforce will prioritise.”
Foundation models rely on significant compute power, and this investment comes on top of around £900 million for a new ‘exascale’ supercomputer and a dedicated AI Research Resource to equip the UK with the processing power it needs to support the next generation of AI innovation.
The taskforce will aim to ensure the budget for compute is strategically invested to prioritise and strengthen the UK’s capability in foundation models.
Sir Adrian Smith, director at the Alan Turing Institute, commented: "Large language models and foundational models can be immensely powerful and have the potential for great benefit, but as with all AI technologies it is vital to understand their limitations and the very real risks associated with them.
“Sovereign AI capability of the kind supported by the Turing and to be taken forward by the taskforce, open to proper scrutiny and developed to the highest ethical standards, stand to have an enormous benefit for public services, society and the economy."