AI tool put to test sifting public views on botox and fillers in UK first
In a pioneering move, an artificial intelligence (AI) tool has been deployed to analyse public responses concerning botox and lip fillers, marking the first such use of this technology within the UK government. Officials hailed the trial as a potential game changer for future public consultations, traditionally considered laborious and resource-intensive.
The AI system, known as “Consult”, was tasked with sorting through around 2,000 submissions gathered during a Scottish Government consultation on the regulation of non-surgical cosmetic procedures. These procedures, including botox and dermal fillers, have seen a surge in popularity over recent years, sparking debates around safety, accessibility, and regulation.
Civil servants involved in the trial reported that the AI-generated results were “nearly identical” to those compiled by human analysts performing the same task manually. The tool successfully identified key themes, categorised responses, and presented summaries that closely matched human outputs.
Consult is the first in a suite of AI-powered tools being developed for use across the UK government under the umbrella moniker “Humphrey”. This is a tongue-in-cheek nod to Sir Humphrey Appleby, the shrewd and manipulative senior civil servant from the classic 1980s satirical sitcom Yes, Minister, which lampooned bureaucratic inefficiency in Whitehall.
According to officials, the deployment of Consult could ultimately free up valuable time for civil servants by automating some of the most time-consuming aspects of policy consultations. Initial projections estimate that widespread adoption of such tools across the public sector could save taxpayers as much as £20 million, with further savings expected if the technology is rolled out more broadly.
Technology Secretary Peter Kyle welcomed the success of the trial, announcing that Consult would be integrated into government processes “very soon”. “After demonstrating such promising results, Humphrey will help us cut the costs of governing and make it easier to collect and comprehensively review what experts and the public are telling us on a range of crucial issues,” he said in a statement.
The government has ambitious hopes that the expanded use of AI technologies like Consult could contribute towards a projected £45 billion in savings across the public sector.
Despite the encouraging results, officials have stressed that Consult remains in its trial phase, and further evaluations are planned before any final decision is made on wider deployment. Importantly, the government has pledged that there will always be “humans in the loop” to oversee the AI’s work, ensuring accuracy and accountability.
Concerns over the potential risks of AI systems, such as the issue of “hallucination”—where AI tools generate false or misleading information—were addressed during the trial. Officials noted that Consult was given a narrowly defined task, reducing the likelihood of such errors occurring.
More Read
Additionally, concerns around AI bias, a known risk associated with large language models which can replicate prejudices found in human-generated training data, were also scrutinised. Interestingly, experts involved in the project found that Consult actually reduced the risk of bias by removing the influence of individual human analysts, who might unconsciously project their own assumptions onto the data.
Moreover, Consult was tested for its robustness in handling spelling mistakes and other textual errors commonly found in public submissions, a capability deemed essential for ensuring inclusive and accurate analysis.
However, limitations remain. At present, Consult only processes responses written in English, meaning that submissions in other UK languages such as Welsh, Scots Gaelic, or Urdu would need to be translated before analysis—a factor the government acknowledged as a gap to be addressed in future iterations.
While the initial trial has been lauded as a success, the government appears cautious but optimistic. Should further testing confirm the AI’s reliability, tools like Consult could become a standard feature in the UK government’s increasingly digital toolkit for public engagement.
