A former director of data science at the UK prime minister's office has told MPs that people working with data in government are not typically technical and would be unlikely to get a similar job in the private sector.
In a hearing designed to illuminate the challenges facing the Department for Science, Innovation and Technology (DSIT) as it strives to become the digital centre for government, MPs quizzed Laura Gilbert, head of AI for Government, at the Ellison Institute and former director of data science at 10 Downing Street, the prime ministers' office.
Members of the House of Common's Science, Innovation and Technology Committee wanted to know about the performance of the Government Digital Service, which in January was moved from the Cabinet Office to DSIT and merged with Central Digital and Data Office (CDDO), the Incubator for AI (i.AI).
Gilbert, a particle physicist who has worked in a number of tech industry roles, said one of the challenges was understanding the level of tech skills in the civil service in central government.
"In the GDS there are a lot a lot of technical skills, and there are pockets of real excellence throughout the system, but there are a lot of people that I come across in government… in probably quite senior digital data roles... who I don't consider are technologists or data people. I wouldn't hire them.
"It's very difficult to hire technologists well and the way the civil service hires [them] is not suitable for this sort of purpose. It's really hit and miss," she said.
"It's very easy to come in with a CV that has all the buzzwords, the jargon and programming languages, and the system doesn't have a way to hire that actually assures [the government] that people do their jobs."
In January, the administration launched a Blueprint for Modern Digital Government, promising that DSIT would work across central government departments to "join up public services" and introduce a training program to help civil service technologists become "AI engineers." It also promised a new package of AI tools, which it nicknames Humphrey. The program was built after a report which suggested central government is missing out on a potential £45 billion ($55 billion) in productivity savings through old or poor use of technology.
Yet Gilbert said the commitment to data integration across the public sector needs to be made for the long term. She said around £250 million ($323 million) was spent and "five separate data exchange projects or initiatives... have come out of GDS or central government as a whole in the last 10 years."
"I'm assuming the Blueprint is signalling another one," she said. "These things are very much long term investments, they take decades."
Earlier this week, a report from Parliament's public spending watchdog warned that legacy IT systems risked hampering the government's attempts to adopt AI. The Public Accounts Committee found that 21 out of the 72 highest-risk legacy systems in government have not been awarded the remediation funding promised in the public sector's 2022-25 Roadmap for Digital and Data.
Speaking to the Commons committee, Richard Pope, a former GDS product manager and currently director of strategy and design business Richard Pope and Partners, told MPs the level of "unmaintained technology" in government and the public sector is "just not good enough at the moment." ®