The government has bold plans to utilise artificial intelligence to ‘boost the economy, provide jobs, and improve people’s everyday lives’.
The UK is already a hub for artificial intelligence in Europe, having twice the number of AI companies compared to any other European nation, with London being considered the AI capital of Europe. The government is actively encouraging the sector, through funding and initiatives to attract talent, as well as implement AI within more of its outputs.
As highlighted in our ‘Geodata trends for 2025’ blog, the government’s ‘AI Opportunities Action Plan’ dedicates an entire section to ‘unlocking data assets in the public and private sector’. As part of this, the government wants to create a ‘National Data Library’ which will strategically collect, curate and make public and private sector data available, with the aim of fostering innovation, research and development within the UK.
An area where AI is already being used to enhance public data is medical research. Within the NHS, artificial intelligence is being used to identify patients for targeted screening to improve early detection of cancer, for example through the Lynch syndrome study.
The Action Plan also highlights how cultural data from bodies such as the National Archives and BBC could be copyright-cleared for British media asset training data, which can then be licensed internationally at scale.
Key within this is using AI to cleanse, collect and curate datasets using both public and private sector knowledge, to improve both innovation and also public sector efficiencies. However, how this move towards AI within the public sector will be received by the UK public is yet to be seen.
The latest Public Opinions and Social Trends survey carried out by the Office for National Statistics reports that 42% of people would not ‘trust any public service to complete some of their tasks using AI’, and only 30% agree or strongly agree with the statement ‘I trust the government to use AI to complete some of its tasks’. Attitudes towards government’s use of AI varies across the demographics, with women and older people being more opposed than both men and younger generations.
The ONS, who carried out the survey, is in itself part of this shifting utilisation of AI for data processing. They have identified over 20 opportunities across the organisation to implement AI, from data processing, preparation and quality assurance, as well as internal operational processes and external communications. In addition to these productivity and efficiency opportunities, the ONS has suggested that the presentation and access of ONS data by large language models may help ensure trust in, and the quality of, statistics and analysis.
Whilst these opportunities are being explored, they have expressed reservation about full implementation, and have developed nine ONS AI Principles to guide their policies, with the UK Statistics Authority implementing similar limitations.
AI has its limitations, such as generating fake content (hallucinations) and algorithmic biases. These ethical issues must be carefully considered when integrating public data into AI systems and models. However, addressing these challenges should not hinder the potential to make data more reliable and accessible for everyone.