\t\t\t\t\tSign up to see the future, today<\/p>\n
Can\u2019t-miss innovations from the bleeding edge of science and tech<\/p>\n
Doomsaying about the coming AI apocalypse is no longer relegated to outbursts from a few tech billionaires<\/a>.<\/p>\n Now, the major American banks whose policies dictate the flow of capital throughout the world are starting to worry about rogue AI, preparing contingency plans for the \u201cday after\u201d a theoretical AI-driven catastrophe.<\/p>\n That\u2019s according to new reporting from Axios<\/a>, which notes that banks are preparing for the worst as they experience a huge rise in AI agent activity on behalf of individual customers.<\/p>\n The main worry seems to be an influx of financial activity driven by autonomous AI agents, which have been authorized to access bank accounts by their human users. In some cases, as Axios noted previously<\/a>, those agents are making moves that haven\u2019t explicitly been prompted by humans \u2014\u00a0a few even going so far as to contact call centers on behalf of their users, without being instructed to do so.<\/p>\n That\u2019s become a major liability for top banks like Morgan Stanley, Bank of America, JPMorgan, and Wells Fargo, which are scrambling to implement security procedures and policy updates to get ahead of the AI swarm.<\/p>\n Failing to stay on top of emerging risks, some experts say, could result in financial catastrophe. As chief economist at the Apollo wealth management firm Torsten Slok explained in a late September brief<\/a>, AI agents operating under instructions to \u201coptimize the return\u201d on their users\u2019 money could mean banks \u201close a large share of the cheap deposits they rely on to make loans, which would be a problem for the entire financial system.\u201d<\/p>\n As skeevy as that admission is, there\u2019s plenty of truth in the fact that our for-profit banking system is simply not designed<\/a> to handle that volume of consumer-friendly optimization. A casino whose slot machines paid out to every player wouldn\u2019t stay in business very long, after all.<\/p>\n Other finance gurus have pointed to the fact that AI agents might simply go haywire,<\/a> regardless of what they\u2019ve been instructed to do.<\/p>\n \u201cYou\u2019re assuming they\u2019re going to act properly on your behalf and they may not,\u201d JPMorgan CEO Jamie Dimon told<\/a> a room full of economists and financial experts in Chicago earlier this month. \u201cYou can\u2019t stop it. You just have to, like, manage the hell out of it.\u201d<\/p>\n