SerpApi.Org
Hackers could use cheap Chinese AI to launch devastating cyberattacks on the world’s financial system, bankers have warned.
The Bank for International Settlements (BIS) said that AI attacks could put financial stability at risk by making it significantly easier to find security flaws.
The BIS, which is known as the central bank of central bankers, said that while it would cost thousands of dollars to carry out an attack with some US systems such as Claude’s Mythos, some models would cost hackers between $50 and $100 (£37-£74).
It pointed to Grok, run by Elon Musk’s xAI, as well as China’s DeepSeek as examples of cheaper systems, adding: “The stakes for the financial system are especially high.”
There have been rising concerns in recent months that Chinese-built open-source systems, which can be tweaked by users, are dominating the cheaper end of the market and becoming increasingly sophisticated.
The BIS said: “As these costs fall, it becomes easier for less sophisticated actors to mount attacks.” It warned that the financial system was “an attractive target for actors seeking to inflict large-scale economic damage”.
In recent days, Chinese AI start-up Moonshot has released Kimi K3, a model it claims rivals the top systems from Anthropic and OpenAI.
Chinese tech giant Alibaba released a new version of its own Qwen system on Sunday. The company says it is second only to Anthropic’s top model.
In a bulletin on AI risk, the BIS said: “The financial system is an obvious place of concern for information technology vulnerabilities.
“Banks, payment systems and other market infrastructures are among the most heavily targeted and most densely interconnected parts of the economy.
“A step change in attackers’ capabilities could therefore have consequences that extend well beyond any single institution and bear directly on financial stability.”
The organisation called for co-operation between banks and national security agencies to boost security, warning that cyber crime was “unusually well suited for frontier AI tools”.
Anthropic has offered access to its Mythos model to a handful of approved organisations under a programme known as Project Glasswing, though Donald Trump recently intervened to prevent the company from offering it to non-US companies on national security grounds.
The BIS, which is owned by the world’s central banks and works to foster financial stability, has also warned that the AI boom could cause a financial crash because of the amounts of debt being used to finance new data centres.
