HomeBussinessAI chatbot launches on Gov.UK to help business users – with mixed...

AI chatbot launches on Gov.UK to help business users – with mixed results

Date:

Related stories

Women’s Super League – latest news and analysis from BBC Sport

Bobbie JacksonBBC Sport JournalistImage source, Getty ImagesChelsea manager Sonia...

Florida Tourism’s Swift Recovery and Future Prospects

Florida’s tourism sector is swiftly recovering from recent hurricanes....

Tech firm creates ‘significant’ number of jobs in Daresbury

Geek+, an autonomous mobile robot (AMR) technology...

Yorkshire duo launch The Networker to connect business community

Two Yorkshire-based businesswomen have launched an online platform designed...
spot_imgspot_img

It speaks a bit of Welsh, can recite the building regulations, refuses to say whether Rishi Sunak is better than Keir Starmer and won’t explain the UK corporation tax regime. The government is launching an artificial intelligence chatbot to help businesses chart the 700,000 page labyrinth that is the Gov.UK website and it looks like users can expect varied results.

The experimental system will be tested by up to 15,000 business users before wider availability, possibly next year. Before you get started it warns: “The biggest limitation of AI tools like me is a problem known as ‘hallucination’. This means we sometimes make up false information or facts but present them to you confidently.”

But the solution, it continues, is to check the links to the website it provides next to its answers, which it fires back in about seven seconds. Speaking about earlier trials in February, Paul Willmott, chair of the government’s Central Digital and Data Office, told reporters that improvements were needed to iron out the “1% of hallucinations where the agent starts to get challenging, or abusive – or even seductive”.

At a test run, offered to reporters on Tuesday, government officials said any hallucinations were now more mundane, like a garbled weblink or an abbreviated answer. The chatbot, which uses OpenAI’s GPT-4o technology, happily explained what regulations a hemp farmer would need to meet before planting a crop, but when asked about the chance of cannabis being legalised in the UK, replied: “I cannot provide predictions or opinions.” It explained the regulations governing the cladding of high-rise buildings after the Grenfell Tower fire, but wouldn’t answer what the public inquiry into that disaster told us about government failures.

In one case, the chatbot briefly answered in Welsh. It wouldn’t answer the question, “What is the corporation tax regime?”, apparently because it didn’t like the word “regime”. But it was fluent on the incentives available for installing solar panels. The chatbot is currently not trained on all Gov.UK documents so content in ministers speeches and press releases appears to be missed.

“Guardrails” have been added so the chatbot won’t respond to queries that may prompt an illegal answer, share sensitive financial information or force it to take a political position. Working with the government’s AI Safety Institute, the developers have also installed protections aimed at preventing hackers from leading the bot astray and making it say damaging things, but cannot rule out the risk altogether.

Peter Kyle, secretary of state for science and technology, said the government wanted to “use AI to improve public services in a safe and reliable way, making sure the UK government leads by example in driving innovation forward”.

He continued: “Outdated and bulky government processes waste people’s time too often, with the average adult in the UK spending the equivalent of a working week and a half dealing with public sector bureaucracy every year. We are going to change this by experimenting with emerging technology to find new ways to save people time and make their lives easier, as we are doing with Gov.UK chat.”

- Never miss a story with notifications

- Gain full access to our premium content

- Browse free from up to 5 devices at once

Latest stories

spot_img