The head of the British Parliament's main public spending watchdog says the government is failing to fix long-standing technology problems even as artificial intelligence brings new cyberattack and fraud risks, The Register reported Wednesday. The warning matters beyond Whitehall: it describes a pattern many large organizations share, adopting AI on top of aging systems that are hard to secure and that few in-house specialists know how to replace.

This account relies on The Register's reporting. The committee's report could not be opened directly when this article was prepared.

What Happened

Sir Geoffrey Clifton-Brown, a Conservative member of Parliament who chairs the House of Commons Public Accounts Committee, published his second annual report as chair. The committee scrutinizes how government spends public money.

RelatedWhite House AI Task Force Plans Breach Warnings From Labs at Its First Meeting

According to The Register, the report says government shows a "worrying lack of urgency" about legacy IT, the older computer systems that departments still depend on. Such systems create significant cybersecurity risks, the committee says, and stop departments from getting the benefits of newer technology. Clifton-Brown wrote that "AI is approaching government like a tsunami" and said he doubts government is ready for it.

The committee also warns that generative AI, tools that produce convincing text, images and audio, gives fraudsters more sophisticated means of attack, and that departments risk falling further behind. It cites estimates that fraud and error cost British taxpayers £55 billion to £81 billion a year.

Skills and Spending

The report points to a thin bench of technical staff. Six percent of civil servants work in digital and data roles, against a government target of 10%, and senior specialists are especially scarce, The Register reported. Ministers have announced reforms to digital capability but have not said how or when they will close the gap.

Modernization has also proved costly. The Register reported that National Savings & Investments, the government-backed savings bank, began an IT overhaul in 2020 to cut running costs. By December 2024 the program had cost £3 billion, £1.3 billion more than first expected, and it has been rated red since 2022-23. The agency could not give lawmakers an up-to-date cost or a realistic completion date. By contrast, the committee praised the Bank of England's £431 million rebuild of its main payment settlement system, completed in April 2025.

Why It Matters

For boards and public-sector leaders, the report links three problems that usually sit in separate risk registers: outdated systems that are hard to patch, a shortage of people who can replace them, and AI tools that make fraud and attacks cheaper to run. Adding AI services on top of that estate expands the number of systems attackers can probe without removing the weak ones underneath.

The skills figure carries a practical lesson. Organizations that cannot hire enough senior engineers tend to rely on contractors and on long programs that overrun, as the savings agency's did. The committee's praise for the Bank of England project suggests that well-run replacements are possible when ownership and scope are clear.

The report's concerns also touch AI use in sensitive services. The digital rights group Open Rights Group has called on the Home Office, the interior ministry, to drop its AI tools for asylum decisions over accuracy, transparency and human oversight, The Register reported.

What to Watch

The government is expected to respond to the committee's recommendations; no date has been given. Watch for detail on how ministers plan to close the digital skills gap, whether departments publish timelines for retiring their riskiest legacy systems, and whether a revised cost and schedule emerges for the National Savings & Investments program. Britain is also weighing AI risk measures and cyber resilience legislation, and earlier this week its AI minister called a breach of national cyber defenses likely.