Bank of England Chief Warns New AI Models Threaten Global Financial Stability
- Reference: 0185325556
- News link: https://news.slashdot.org/story/26/08/31/1745218/bank-of-england-chief-warns-new-ai-models-threaten-global-financial-stability
- Source link:
> Writing in his capacity as chair of the Financial Stability Board, an international body that coordinates policy and makes recommendations to national authorities, Bailey identified the potential impact of frontier AI -- which refers to the most advanced AI models -- on cyber risk as "the most immediate concern" for the financial system.
>
> "Frontier AI may have the ability materially to alter the speed, scale and economics of cyber risk, which could undermine market confidence system-wide, especially due to highly concentrated third-party service providers," Bailey said. "Recent developments have also highlighted to me that many jurisdictions do not have the protocols in place to manage the development, release, and deployment of advanced frontier AI models, heightening risks for the financial sector and beyond," he added.
>
> [...] Financial institutions and technology providers will need to improve vulnerability management, response and recovery capabilities -- "and prepare for more severe scenarios involving simultaneous disruption across multiple firms or shared technology dependencies," Bailey said. Alongside new AI models, Bailey cited "fragilities" in sovereign debt markets, the growing use of debt by investors in equity markets and stretched asset valuations, particularly AI-related investments, as among his concerns.
[1] https://www.cnbc.com/2026/08/31/bailey-frontier-ai-financial-stability-risk.html
[2] https://www.fsb.org/2026/08/fsb-chairs-letter-to-g20-finance-ministers-and-central-bank-governors-august-2026/
could? (Score:2)
I would say that they already have, and will continue to do so.
Attacks? (Score:3)
Robinhood may become a thing of the past. But I'm making oodles of money with a broker who takes my orders over the phone.
Never mind [1]paperclips [wikipedia.org], what happens when AI becomes self serving and starts to divert all capital into AI data centers?
[1] https://en.wikipedia.org/wiki/Instrumental_convergence#Riemann_hypothesis_catastrophe
Re: (Score:2)
> what happens when AI becomes self serving and starts to divert all capital into AI data centers?
Nothing changes. It’s already happening. If only I had purchased 64gb of rdimm with my server, I could have sold it to pay off my mortgage.
And that's a bad thing? (Score:2)
I feel like I'm being threatened with a good time.
Amateur Hour in IT Security is over (Score:2)
If you run software where security weaknesses typically can be exploited, you will get attacked. Only a matter of time and the attackers will soon have their own local models to do it (or already have them). Any patch will immediately lead to the creation of attack code. You have hours or less to patch. Unless you have effective defense-in-depth and security patches are for when one of several effective security mechanisms has a problem but the others are still fine, i.e. the vulnerability is non-exploitabl
Good News! (Score:2)
At least if you believe in the Bank of England rumor:
> According to this image, Afghanistan, Iraq, Sudan, Libya, Cuba, North Korea, Iran, Syria, and Venezuela were the only countries without a central bank controlled by the family's [Rothschild ] well-known financial and banking groups that year.
but it is a false rumor: [1]https://factcheck.afp.com/doc.... [afp.com]
[1] https://factcheck.afp.com/doc.afp.com.36RR8D7
Of all people, they have options. (Score:2)
If ANYONE can finance good security, it's a bank. Among banks, central banks are in better positions than other banks to properly fund security. At this point, they should be working on finalizing their own extremely well-tested OS build. Seriously, their code should be rock-solid and "old" instead of chasing the latest computing fad. Also, if they are using close-source stuff like Windows then they aren't even pretending to be serious about security.
If this statement proves anything then it is that for dec
Re: (Score:2)
I think the problem is that there's a conflict between good security and ease of remote use. They've been pushing ease of remote use with "adequate" security. Now that security is no longer adequate, but they don't want all the customers to be angry with them.
Not their primary fear (Score:2)
What it is, is people being able to ask plain english simple questions to understand how something works, the value of it, how much will cost them over x years, and what would be better options from the complex information that can confuse people not familiar with finances and banking getting straight simple answers.
In short, they're worry that regular people would become educated to the bad deals that make the banks so much money and make wiser financial decisions, which will result in less money for them,
In future AI news ... (Score:2)
The movies The Matrix and The Terminator are regarded as documentaries.
Re: In future AI news ... (Score:3)
Itâ(TM)s only marketing.