Government security experts have warned that criminals could be capable of using artificial intelligence to break into bank accounts within just three to five months, raising the prospect of what one Cabinet Office insider called an "AI Armageddon". The concern, relayed to the Mail on Sunday by officials linked to Britain's AI Security Institute (AISI), stems largely from Chinese AI labs releasing powerful models with weak or easily removed safety guardrails, which could let hackers worldwide bypass strong passwords and steal financial data far sooner than previously expected.

The Financial Stability Institute warned last Wednesday that banks are struggling to keep pace with such rapidly advancing technology, and in some cases may have only hours to block an intrusion once it begins; it urged financial institutions to strengthen their defences. UK ministers are reportedly considering more "targeted interventions" to manage the national security risk without harming growth or innovation. The warning comes amid wider AI safety tensions, including Anthropic barring Britain's safety institute from testing its newest model, the firm's disclosure that it stopped five attempted bioweapon misuse cases this year, and a departing Anthropic researcher's claim of a greater than 10% chance AI could cause human extinction within a decade — prompting campaigners and MPs to call for the government to take AI risks more seriously.